WO2007062859A1 - A method for predictive detection of deterioration in a wireless network channel - Google Patents

A method for predictive detection of deterioration in a wireless network channel Download PDF

Info

Publication number
WO2007062859A1
WO2007062859A1 PCT/EP2006/011548 EP2006011548W WO2007062859A1 WO 2007062859 A1 WO2007062859 A1 WO 2007062859A1 EP 2006011548 W EP2006011548 W EP 2006011548W WO 2007062859 A1 WO2007062859 A1 WO 2007062859A1
Authority
WO
WIPO (PCT)
Prior art keywords
throughput
sampling
wireless network
average
signal strength
Prior art date
Application number
PCT/EP2006/011548
Other languages
French (fr)
Inventor
Davide Lenzarini
Mirko Kulig
François SCHWITTER
Alessandro Puiatti
Silvia Giordano
Original Assignee
Forward Information Technologies S.A.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Forward Information Technologies S.A. filed Critical Forward Information Technologies S.A.
Publication of WO2007062859A1 publication Critical patent/WO2007062859A1/en

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B17/00—Monitoring; Testing
    • H04B17/30—Monitoring; Testing of propagation channels
    • H04B17/309—Measuring or estimating channel quality parameters
    • H04B17/318—Received signal strength
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B17/00—Monitoring; Testing
    • H04B17/20—Monitoring; Testing of receivers
    • H04B17/23—Indication means, e.g. displays, alarms, audible means
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W16/00—Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
    • H04W16/22—Traffic simulation tools or models

Definitions

  • TITLE A method for predictive detection of deterioration in a wireless network channel.
  • the present invention relates to a method for predictive detection of a deterioration in a wireless network channel, whereto an electronic mobile device is connected through at least one wireless network interface, the electronic mobile device comprising a software application for transmitting/ receiving a Constant Bit rate traffic to/from the wireless network channel.
  • the present invention relates to a method of the type described above, wherein a detection phase issues an alert signal to a final user thereby signaling the deterioration of the wireless network channel, before the lost of the communication.
  • an electronic mobile device intended to communicate over a wireless network channel comprises at least one physical wireless network interface, for connecting-to and transmitting- over the wireless network channel.
  • the electronic mobile device may be a laptop, a PDA, a smart-phone or a similar device intended to communicate over a wireless network channel of the type Bluetooth, Wi-Fi, GPRS, UMTS, EDGE, HSDPA, IXRTT, CDMA2000 or others.
  • the electronic mobile device also comprises one or more software applications, implementing corresponding services for a user: for example an application for a voice service, an application for video streaming or for ftp uploading/ downloading service from/ to a remote server.
  • software applications implementing corresponding services for a user: for example an application for a voice service, an application for video streaming or for ftp uploading/ downloading service from/ to a remote server.
  • the services have different requirements in term of channel bandwidth , bit error rate, end-to-end delay, jitter and availability, since they generate different types of traffic, such as:
  • CBR Constant Bit Rate
  • ABR Available Bit Rate
  • VBR Variable Bit Rate
  • Constant Bit Rate traffic is usually generated by software applications for voice service, video streaming, ftp uploading/ downloading service or the like, which are assumed to be always supported by a traffic constantly and at a given rate.
  • end-to-end delay i.e. the transfer delay from two devices communicating over the wireless network channel
  • variability in the end-to-end delay also referred to as "jitter”
  • packet loss rate the fraction of packets of data lost
  • Constant Bit Rate traffic may be distinguished in real time Constant Bit Rate traffic, wherein the delay constraints are strict, for example to render the video/ audio quality acceptable to a final user, and not real time Constant Bit Rate traffic, having a less strict delay constraints, for example in case of ftp service, since the reading of a file may be delayed when all its data are uploaded/downloaded.
  • the wireless network channel should be carefully monitored.
  • Seamless handover methods are known to manage a switch from a first to a second connection between a client or software application, running on an electronic device, and a server application, running on a remote server, without interrupting a data transfer in progress.
  • an electronic mobile device may transmits and/ or receives data from a software application, running on it, and a server application, running on a remote server, connected to the electronic mobile device by means of a first wireless network channel.
  • a user may then switch the electronic mobile device to connect to a second wireless network channel.
  • a seamless handover method is mainly aimed to manage such a transition between the two wireless network channels in a transparent way for the user, for example lowering for a while the transmission rate of the data, without interrupting the data transfer.
  • Misikangas requires very high resources to the electronic mobile device while it does not guarantee a response time for the prediction suitable to audio and video real time services.
  • the problem at the basis of the present invention is that of providing a method for detecting the deterioration of a wireless network channel, whereto an electronic device is connected through a wireless network interface and comprises a software application which produces a substantially Constant Bit rate traffic, useful for example in case of video or audio communications with one or more remote servers, the method providing predicting information for high OSI layer software applications before the communication becomes impossible, without requiring to the electronic mobile device computational intensive operations nor additional information received from said servers, nor from additional access points or devices belonging to the wireless network, overcoming all the problems that currently affect the detection of the deterioration in wireless channel according to the prior art.
  • the solution idea on which the present invention is based is that of providing a method for predictive detection of a wireless network channel degradation, evaluating widely accessible parameters and functions related to a high OSI layer software application that is associated to a Constant Bit rate traffic, the method taking into account the average throughput transmitted /received by an electronic mobile device to/from the wireless network channel.
  • a method for detecting a deterioration in a wireless network channel whereto an electronic mobile device is connected through at least one wireless network interface, the electronic mobile device comprising a software application, for transmitting/ receiving a substantially Constant Bit Rate traffic to/from the wireless network channel, the method comprising
  • a signal strength sampling phase for sampling an average strength of a signal, received by the at least one wireless network interface on the wireless network channel
  • a detection phase for detecting when the average strength reaches a predefined threshold value
  • a bound sampling phase defining a lower and an upper bound of a throughput threshold interval by sampling a throughput generated by the software application over the wireless network channel;
  • a throughput sampling phase for sampling an average of the throughput
  • the detecting phase comparing said average of the throughput with respect to said lower and upper bounds of the throughput threshold interval and issuing an alert signal to a final user thereby signaling said deterioration of the wireless network channel, before the lost of the communication .
  • a detection of the deterioration of the wireless network channel, whereto the electronic mobile device is connected is determined before that the substantially Constant Bit Rate traffic produced by the software application is delayed to the user, and thus before rendering for example a video or audio communication unusable, such a detection not requiring computational intensive operations to the electronic mobile device nor additional resources external to it.
  • Figure 1 schematically shows, in a block diagram, the main phases of the method for predictive detection of deterioration in a wireless network channel according to the invention.
  • Figure 2 schematically shows, in a block sub-diagram 2, one phase of the method according to the invention.
  • Figure 3 schematically shows, in a block sub-diagram 3, another phase of the method according to the invention.
  • Figure 4 schematically shows, in a block sub-diagram 4, a further phase of the method according to the invention.
  • Figure 5 schematically shows, in a block sub-diagram 5, yet another phase of the method according to the invention.
  • a method for predictive detection of degradation in a wireless network channel, over which an electronic mobile device 1 communicates is schematically indicated with block diagram 60.
  • electronic mobile device 1 is referred to devices of the type comprising one or more physical wireless network interfaces 12, for supporting corresponding one or more network communications over wireless network channels.
  • the electronic mobile device 1 could be a laptop, a PDA, a smart-phones or a similar devices.
  • some example of physical wireless network interfaces 12 are interfaces for an optical network, like infrared networks, for Wi-Fi networks according to the IEEE 802.11 standard, for Bluetooth network according to the IEEE 802.15.1 standard, for ZigBee networks according to IEEE 802.15.4 standard, for GPRS networks, for EDGE networks, for IXRTT networks, for CDMA2000 networks, for UMTS networks, for HSDPA networks, for WiMAX networks according to the IEEE 802.16 standard and the like.
  • an optical network like infrared networks, for Wi-Fi networks according to the IEEE 802.11 standard, for Bluetooth network according to the IEEE 802.15.1 standard, for ZigBee networks according to IEEE 802.15.4 standard, for GPRS networks, for EDGE networks, for IXRTT networks, for CDMA2000 networks, for UMTS networks, for HSDPA networks, for WiMAX networks according to the IEEE 802.16 standard and the like.
  • the electronic mobile device 1 comprises a software application 11 that, in execution, generates a substantially Constant Bit Rate traffic.
  • a software application 11 that, in execution, generates a substantially Constant Bit Rate traffic.
  • an audio and/ or video application having stringent time and bandwidth requirements, produces a continuous and almost Constant Bit Rate traffic of real time data.
  • the software application 11 may also be a real time voice call application, a video call application, a video-on- demand application, a mobile TV applications, a mobile radio application or similar.
  • the electronic mobile device 1 also comprises a network communication module, for transmitting/ receiving a number of frames at Constant-Bit Rate traffic over the wireless communication channel, through the above mentioned interface 12.
  • a detection phase 2 for detecting the wireless channel degradation may be activated on the electronic mobile device 1 in order to preventively determine the availability of the channel. More particularly, the detection phase may be recalled by a user, for example through a keyboard on the electronic mobile device 1, or by a system user, according to predefined conditions; the human user and the system user are schematically indicated in figures with numeral reference 7 and also indicated as end-consumer.
  • the predictive detection can be used by an end-consumer 7 to bring into being all the necessary pro-active activities to minimize the quality of service reduction of the software application 11, i.e. the throughput reduction and the user experience worsening. Said pro-active activities require the application or its user intervention to react to the different and worse wireless channel conditions.
  • a seamless handover system can, in fact, use the predictive detection of the wireless channel degradation to start the selection of a new suitable wireless channel, if present, and connect to it automatically, if possible, and transparently let the application continue to work as before minimizing the throughput reduction and the user experience worsening.
  • signal strength sampling phase 3 intended to determine the average signal strength on the basis of samples of the signal strength retrieved from the at least one wireless network interface 12 with a predetermined time interval, hereinafter indicated as signal strength sampling time interval;
  • a bound sampling phase 4 intended to determine an upper and a lower bound of a throughput threshold interval defined on the basis of the average of the throughput generated by the software application 11 over the wireless communication channel and sampled at a low sampling time interval;
  • a throughput sampling phase 5 intended to determine the average of the throughput generated by the software application 11 over the wireless communication channel and sampled at a high sampling time interval;
  • timing phase 6 triggering the execution of the above mentioned phases.
  • the above described phases are to be considered as tasks or threads which are contemporary performed. More particularly, the detection phase 2, the signal strength sampling phase 3, the bound sampling phase 4 and the timing phase 6 are in execution till the signal strength value is above a warning threshold value, while the detection phase 2, the signal strength sampling phase 3, the throughput sampling phase 5 and the timing phase 6 are executed when the signal strength value overcomes such a warning threshold value, as will be more clearly described in the following lines.
  • the signal strength sampling phase 3 is recalled by the detection phase 2 in order to calculate the average value of the signal strength on the basis of a plurality of values or samples stored in a first or signal strength buffer 31.
  • the signal strength sampling phase 3 initializes the signal strength buffer 31 with a predetermined value, indicated as signal strength warning threshold value, for example stored in a non volatile memory portion of the electronic mobile device 1.
  • the signal strength warning threshold value may be configurable, for example on the basis of statistics related to the specific wireless technology used for communication by the electronic mobile device 1.
  • the signal strength buffer 31 comprises a plurality of memory locations for storing a plurality of corresponding signal strength samples; more particularly, the signal strength buffer 31 is a rotational buffer, per se known in the art.
  • the signal strength buffer 31 could be considered as a linear buffer wherein the samples of the signal strength are stored from left to right inside the signal buffer 31 itself.
  • a new sample may be stored in the leftmost position with the right shifting of all the previously stored data and the discarding of the rightmost one as it stores the oldest and less updated sample.
  • the signal strength sampling phase 3 stores data also in a second or average signal strength buffer 32, for example similar to the signal strength buffer 31 and in particular a rotational one. The first time it is activated, the signal strength sampling phase 3 initializes also the average signal strength buffer 32 with the signal strength warning threshold value.
  • the signal strength sampling phase 3 waits for a trigger from the timing phase 6 in order to retrieve and evaluate new signal strength samples.
  • the signal strength sampling phase 3 retrieves from the at least one wireless network interface 12 at least a new sample and inserts it in the signal strength buffer 31.
  • the first time such sample may be stored in the leftmost position of the signal strength buffer 31 , the second time in the following position and so on.
  • the signal strength sampling phase 3 calculates an average value of the values i.e. the samples stored in the signal strength buffer 31 and stores such average value in the average signal strength buffer 32. Also for the average signal strength buffer 32, the first time, such average value may be stored in the leftmost position of the average signal strength buffer 32 itself, the second time in the following position and so on.
  • Such a signal strength sampling phase 3 performs the above described operations and samplings, each time it is triggered by the timing phase 6.
  • the bound sampling phase 4 is recalled by the detection phase 2 in order to calculate a throughput warning threshold interval, on the basis of a plurality of values stored in a third or bound buffer 41 and determined by the throughput associated to the software application 11. More particularly, the throughput warning threshold interval is calculated with boundaries that are fractions of the average of values stored in the bound buffer 41 , as described hereafter.
  • such a bound sampling phase 4 calculates the throughput warning threshold interval on the basis of the average throughput value when the software application 11 normally works (for instance, the throughput warning threshold interval currently used is between -9% and -15% of the average throughput value).
  • the bound sampling phase 4 initializes the bound buffer 41 with a predetermined value, for example "0" in all the locations or a value stored in a non volatile memory portion of the electronic mobile device 1; also this value may be configurable.
  • the bound buffer 41 comprises a plurality of memory locations for storing a plurality of corresponding values of the throughput; more particularly, the bound buffer 41 is a rotational buffer, for example of the type described with reference to the signal strength buffer 31.
  • the bound sampling phase 4 waits to be triggered by the timing phase 6; when triggered, it samples the throughput generated by the software application 11 and insert it in the bound buffer 41.
  • the bound sampling phase 4 when the bound sampling phase 4 is triggered by the timing phase 6, it may return in a waiting state if the detection phase 2 does not authorize it to proceed in the sampling of frames. This case, as it will be clear later, happens when the detection phase 2 considers the signal strength sufficiently low to require deeper analysis of the throughput.
  • the bound sampling phase 4 calculates boundaries of the throughput warning threshold interval on the basis of fractions of the average of values stored in the bound buffer 41. More particularly, the throughput warning threshold interval boundaries may be calculated according to different advantageous formulas and can also have the same value.
  • Such a bound sampling phase 4 performs the above described operations and sampling each time it is triggered by the timing phase 6.
  • the throughput sampling phase 5 is recalled by the detection phase 2 in order to calculate, on the basis of a plurality of values stored in a fourth buffer 51, an average value of the throughput generated by the software application 11 as described hereafter.
  • the throughput sampling phase 5 initializes fourth and fifth buffers, 51 and 52, also indicated as throughput and average throughput buffers, with a predetermined value.
  • a predetermined value may be the upper bound of the throughput warning threshold interval calculated by the bound sampling phase 4.
  • the throughput and average throughput buffers 51 and 52 comprise a plurality of memory locations for storing a plurality of corresponding values.
  • the throughput and average throughput buffers 51 and 52 are rotational buffers, as explained with reference to the signal strength buffer 31.
  • the throughput sampling phase 5 waits to be triggered by the timing phase 6; when triggered, it samples the throughput generated by the software application 11 and insert it in the throughput buffer 51.
  • the bound sampling phase 4 when the bound sampling phase 4 is triggered by the timing phase 6, it may return in a waiting state if the detection phase 2 does not authorize it to proceed in the sampling of frames. This happens when the detection phase 2 considers the signal strength sufficiently high to avoid deeper analysis of the throughput.
  • the throughput sampling phase 5 calculates the average value of the values stored in the throughput buffer 51 and stores such average value in the average throughput buffer 52.
  • the signal strength sampling phase 3 is responsible to sample signal strengths in order to evaluate the average strength of the signal respect to a signal strength threshold;
  • the bound sampling phase 4 is responsible to sample at a low rate the throughput generated by the software application 11, in order to determine a throughput warning threshold interval, comprising a lower and an upper bound;
  • the throughput sampling phase 5 is responsible to sample at a high rate the throughput generated by the software application 11, in order to determine an average value of said throughput, a starting value of said average value being related to the upper bound calculated in bound sampling phase 4, that corresponds to a potential starting point of a deterioration of the wireless network channel.
  • the signal strength sampling phase 3 and the bound sampling phase 4, which are triggered by the timing phase 6, support the detection phase 2 in order to evaluate the degradation of the wireless network channel: when the detection phase 2 considers that the wireless network channel is working normally, as the signal strength is greater than the signal strength warning threshold value, only the signal strength sampling phase 3 and the bound sampling phase 4 are activated.
  • the throughput sampling phase 5 does not work and the bound sampling phase 4 samples at a low rate the throughput in order to provide the throughput warning threshold interval as described: in this case, a warning counter is set to 0 in order to remember that the current state of the wireless network channel is of non deterioration.
  • the detection phase 2 When the detection phase 2 detects a deterioration (on the basis of the signal strength warning threshold value), it requires to the bound sampling phase 4 the last updated throughput warning threshold interval, stops the bound sampling phase 4 and activates the throughput sampling phase 5.
  • the detection phase 2 compares the lower and upper bounds of the last throughput warning threshold interval updated by the bound sampling phase 4 with the average throughput calculated by the throughput sampling phase 5.
  • the detection phase 2 has detected a deterioration of the wireless network channel and issues an alert to the human or system user 7;
  • the detection phase 2 does not issue any alert to the human or system user 7 but increases the warning counter by 1, as an indication that the condition of the wireless network channel is probably worsening (i.e. it is probably deteriorating);
  • the detection phase 2 has detected a deterioration of the wireless network channel and issues an alert to the human or system user 7;
  • the detection phase 2 set the warning counter to 0, as a feedback that the condition of the wireless network channel is returning in acceptable values (i.e. the wireless network channel is not deteriorated).
  • the detection phase 2 may be activated by a human user 7 in order to monitor the wireless network channel and to be alerted if a deterioration occurs.
  • the detection phase 2 may also be activated by a system user 7, for example by a process requiring the monitoring of the deterioration of the wireless network channel (e.g. a seamless handover system) .
  • the timing phase 6 is set so as to activate an execution of the detection phase 2 at predefined events or intervals, for example:
  • the execution time interval in order to avoid unnecessary executions of the detection phase 2 on unmodified data, may be greater than the minimum between the signal strength sampling time interval and the high sampling time interval.
  • the execution time interval may be chosen so that the difference between its duration and the minimum between the signal strength sampling time interval and the high sampling time interval is substantially equal to the time required by the detection phase 2 to complete a single execution.
  • the timing phase 6 may be set by the signal strength sampling phase 3, the bound sampling phase 4 and by the throughput sampling phase 5.
  • the signal strength sampling phase 3 sets the timing phase 6 so that a trigger may activate phase 3 itself in order to sample the signal strength on the wireless network channel; hereinafter such trigger is indicated as signal strength sampling trigger;
  • the bound sampling phase 4 sets the timing phase 6 with a low sampling trigger, so that a trigger may activate phase 4 itself in order to sample the throughput at low rate and to determine the corresponding lower and upper bounds of the throughput warning threshold interval;
  • the throughput sampling phase 5 sets the timing phase 6 with a high sampling trigger, so that a trigger may activate phase 5 itself in order to sample the throughput at a high rate.
  • the bound sampling phase 4 is executed only when the wireless channel is in normal conditions, in other words when the throughput samples fluctuations are not relevant, while the throughput sampling phase 5 is used when the channel conditions are getting worst, when the short time determination of an average throughput negative trend may bring to a predictive detection of the wireless communication channel degradation.
  • the high sampling time interval may be chosen so as to be greater than the time interval between two consecutive packets generated by the Constant Bit Rate traffic of the software application 11, in order to avoid a throughput sample equal to zero and so determining a false prediction of the wireless communication channel degradation.
  • the setup time in the timing phase 6, associated to the detection phase 2 is greater than or equal to the time requested by the signal strength sampling phase 3 and the bound sampling phase 4 to start in order to produce valid average values based on a reasonable number of real sampled data.
  • the detection phase 2 enter a first loop of execution recalling the signal strength sampling phase 3.
  • the signal strength sampling phase 3 returns the last average signal strength calculated on the basis of the signal strength samples, provided by the at least one wireless network interface 12 of the electronic mobile device 1.
  • This average value is compared with a signal strength warning threshold value, that is for example a predetermined value or a configurable value, indicating the signal strength value at which it is required a more intensive analysis on the throughput generated by the software application 11.
  • a signal strength warning threshold value that is for example a predetermined value or a configurable value, indicating the signal strength value at which it is required a more intensive analysis on the throughput generated by the software application 11.
  • the throughput generated by the software application 11 cannot be influenced by the signal strength variations because it is so strong that its strength fluctuations cannot have direct effects on the throughput.
  • the throughput generated by the software application 11 could be heavily influenced by the signal strength variations, for example when the electronic mobile device 1 approaches a border of a wireless cell whereto it is currently connected.
  • an evaluation of physical layer parameters is only used as a trigger for a deeper analysis involving OSI data link and/or OSI network and/or OSI transport and/or OSI application layer parameters collected at a proper sampling rate.
  • the software application 11 when the average signal strength falls down the signal strength warning threshold, a deeper analysis of the throughput as generated by the software application 11 is started. Such a deeper analysis is carried out by using values that are sampled at a higher rate than a normal condition rate, normally used to calculate the reference average throughput.
  • This critical signal strength value is typical for each specific wireless technology since the signal strength has different power and different characteristics from one wireless technology to another. Furthermore, inside a same wireless technology, this critical signal strength value varies depending on the hardware and software of the wireless interfaces, for example depending on the specification followed by the different manufacturers or on the interface sensitivity.
  • the signal strength warning threshold depends only on the wireless technology considered and not on the particular wireless interface used and it can be derived from statistical studies for instance as a fraction of the average critical signal strength thresholds.
  • the timing phase 6 may be reset to trigger the detection phase 2 after a sleep time has elapsed. This sleep time should be greater or equal to the difference between the signal strength sampling time interval and execution time interval.
  • An abnormal condition corresponding to a potential degradation of the wireless network channel, occurs when the last average signal strength is lower than the signal strength warning threshold value.
  • the detection phase 2 requests to the bound sampling phase 4 the last throughput warning threshold interval calculated.
  • This interval has boundaries calculated as fractions of the average low sampling throughput derived from the number of frames generated by the constant-bit rate traffic in the low sampling time interval and provided by an operating system network communication module 13 of the electronic mobile device 1.
  • the detection phase 2 When the detection phase 2 has received the last throughput warning threshold interval, it starts the throughput sampling phase 5, if not already running, and requests to it the last average throughput calculated.
  • This value is derived from the number of frames generated by the constant-bit rate traffic during the high sampling time interval and provided by an operating system network communication module 13 of the electronic mobile device 1.
  • the detection phase 2 detects the degradation of the wireless network channel before the communication over it becomes impossible and signals it to the user 7, for example to an end-consumer.
  • the constant-bit rate throughput generated by the software application 11 must be much greater (at least one order of magnitude) than the eventual average variable-bit rate throughput generated by the mobile device operating system and/ or by the eventual other applications running on the electronic mobile device 1.
  • This eventual variable-bit rate traffic can be considered as noise and this noise has to be much lower than the constant signal represented by the constant-bit rate traffic generated by the software application 11 , in order to detect with reasonable sureness when the throughput starts to fall, before the communication becomes impossible.
  • the present invention also relates to an electronic mobile device 1 for predictive detection of deterioration in a wireless network channel.
  • the electronic mobile device 1 thus comprises:
  • an operating system network communication module 13 for transmitting/ receiving the frames of the Constant-Bit Rate traffic to /from the wireless network channel
  • At least one wireless network interface 12 for connecting the electronic device 1 to the wireless network channel
  • detecting means 2 for alerting a user 7 of the application about said deterioration.
  • the electronic mobile device 1 also comprises the following:
  • signal strength sampling means 3 for sampling an average strength of a signal, received by the at least one wireless network interface 12 on the wireless network channel;
  • bound sampling means 4 defining a lower and an upper bound of a throughput threshold interval by sampling a throughput generated by the software application 11 ;
  • throughput sampling means 5 for sampling an average of the throughput.
  • the detecting means 2 when the average strength of the signal reaches a predefined threshold, raises the predictive detection if the average throughput is out of bound with respect to the lower and upper bounds of the throughput threshold interval.
  • the electronic mobile device 1 also comprises a timer 6 for triggering, at corresponding predefined intervals, executions of
  • the detecting means 2, signal strength sampling means 3, bound sampling means 4 and throughput sampling means 5 set the corresponding predefined intervals in the timer 6.
  • the signal strength means 3 samples a plurality of values of the signal strengths and stores it in a first or signal strength buffer 31, comprising a corresponding plurality of memory locations; averages of the plurality of values are stored in successive memory locations of a second or average signal strength buffer 32.
  • the bound sampling means 4 samples a plurality of throughput values and stores it in a third or bound buffer 41 , comprising a corresponding plurality of memory locations, the lower and upper bounds of the throughput threshold interval being determined on a computation of the plurality of the throughput values.
  • the throughput sampling means 5 samples a plurality of throughput values and stores it in a fourth or throughput buffer 51, comprising a corresponding plurality of memory locations; averages of the plurality of the throughput values being stored in successive memory locations of a fifth or average throughput buffer 52.
  • the detecting means 2, the signal strength sampling means 3, the bound sampling means 4 and the throughput sampling means 5 are implemented through corresponding hardware modules. Alternatively, these means can be part of a more complex apparatus or implemented by software applications.
  • a detection of the deterioration of a wireless network channel, whereto an electronic mobile device is connected through a wireless network interface, producing a substantially Constant Bit Rate traffic is determined in time to alert a user before the connection falls.
  • the detection according to the present invention is particularly advantageous in video or audio communications between the electronic mobile device and one or more remote servers, especially in predicting deterioration in a wireless network channel used by high layer software applications before the communication becomes impossible.
  • the method and device does not require computational intensive operations or additional information to said remote server, or to additional access points or devices belonging to the wireless network, and it is based on the resource normally available in electronic mobile devices.

Abstract

Method for predictive detection of a deterioration in a wireless network channel, whereto an electronic mobile device (1) is connected through at least one wireless network interface (12), the electronic mobile device (1) comprising a software application (11) for transmitting/ receiving a Constant Bit rate traffic to/from the wireless network channel, the method comprising a detection phase (2) for issuing an alert signal to a final user (7) thereby- signaling said deterioration of the wireless network channel, before the lost of the communication. The method includes a signal strength sampling phase (3) for monitoring an average strength of a signal, received by the at least one wireless network interface (12) on the wireless network channel; a bound sampling phase (4) for defining a lower and an upper bound of a throughput threshold interval by sampling a throughput generated by the software application (11); a throughput sampling phase (5) for monitoring an average of the throughput.

Description

TITLE: A method for predictive detection of deterioration in a wireless network channel.
DESCRIPTION
Field of Application
The present invention relates to a method for predictive detection of a deterioration in a wireless network channel, whereto an electronic mobile device is connected through at least one wireless network interface, the electronic mobile device comprising a software application for transmitting/ receiving a Constant Bit rate traffic to/from the wireless network channel.
More particularly, the present invention relates to a method of the type described above, wherein a detection phase issues an alert signal to a final user thereby signaling the deterioration of the wireless network channel, before the lost of the communication.
Prior Art
As it is well known, an electronic mobile device intended to communicate over a wireless network channel comprises at least one physical wireless network interface, for connecting-to and transmitting- over the wireless network channel.
For example, the electronic mobile device may be a laptop, a PDA, a smart-phone or a similar device intended to communicate over a wireless network channel of the type Bluetooth, Wi-Fi, GPRS, UMTS, EDGE, HSDPA, IXRTT, CDMA2000 or others.
According to the wireless network channel whereto the electronic mobile device is intended to connect, it is provided with a specific physical wireless network interface, for example an optical interface, an infrared interface, a Wi-Fi interface according to the IEEE 802.11 standard, a Bluetooth interface complying with the IEEE 802.15.1 standard or other interfaces.
More particularly, the electronic mobile device also comprises one or more software applications, implementing corresponding services for a user: for example an application for a voice service, an application for video streaming or for ftp uploading/ downloading service from/ to a remote server.
The services have different requirements in term of channel bandwidth , bit error rate, end-to-end delay, jitter and availability, since they generate different types of traffic, such as:
- Constant Bit Rate (CBR) traffic,
- Unspecified Bit Rate (UBR) traffic,
- Available Bit Rate (ABR) traffic,
- Variable Bit Rate (VBR) traffic,
- Real time and (not) real time Bit Rate traffic.
More particularly, the Constant Bit Rate traffic is usually generated by software applications for voice service, video streaming, ftp uploading/ downloading service or the like, which are assumed to be always supported by a traffic constantly and at a given rate. In case of Constant Bit Rate traffic, the end-to-end delay (i.e. the transfer delay from two devices communicating over the wireless network channel), the variability in the end-to-end delay (also referred to as "jitter"), and the fraction of packets of data lost (also indicated as "packet loss rate") should be guaranteed to be less than some specified values.
Constant Bit Rate traffic may be distinguished in real time Constant Bit Rate traffic, wherein the delay constraints are strict, for example to render the video/ audio quality acceptable to a final user, and not real time Constant Bit Rate traffic, having a less strict delay constraints, for example in case of ftp service, since the reading of a file may be delayed when all its data are uploaded/downloaded.
Especially in case of services requiring real time Constant Bit Rate traffic, the wireless network channel should be carefully monitored.
In fact, the connection of an electronic mobile device over the wireless network channel may suffer high variability, for example due to a change of the location of the mobile electronic device and to a non constant cover of the wireless network channel in the new location, corresponding to quality fluctuations of the services rendered to the final user and, in the worst case, to suspension of the services.
Seamless handover methods are known to manage a switch from a first to a second connection between a client or software application, running on an electronic device, and a server application, running on a remote server, without interrupting a data transfer in progress.
For example, an electronic mobile device may transmits and/ or receives data from a software application, running on it, and a server application, running on a remote server, connected to the electronic mobile device by means of a first wireless network channel. A user may then switch the electronic mobile device to connect to a second wireless network channel. A seamless handover method is mainly aimed to manage such a transition between the two wireless network channels in a transparent way for the user, for example lowering for a while the transmission rate of the data, without interrupting the data transfer.
Example of seamless handover method are described in the PCT applications N° WO2005/ 076651 in the name of Forward Information Technologies SA, N° WO 02/ 103978 A2 in the name of Swisscom Mobile AG, N0 WO 03/065682 Al in the name of KONINKLI JKE PHILIPS ELECTRONICS N. V., and N° WO 02/43348 Al in the name of Columbitech AB.
Other method are described in the European patent applications N° EP 1 089 495 A2 in the name of Nortel Networks Limited, N° EP 0 998 094 A2 in the name of Nokia Mobile Phones LTD and N° EP 1 322 089 A2 in the name of Assimakopoulos Theodoros.
In such documents, in order to determine if and when a seamless handover between different wireless network channels is needed, some parameters related to a current connection are collected and compared with predefined critical threshold values thereof.
Even if such methods face the problem of switching between wireless network channels, they fails to guarantee the quality of the service in case of real time Constant Bit Rate traffic, since they all act when the state of the connection is already compromised rather than predict the degradation of the channel. Furthermore, if the switching is delayed, the quality of service can be reduced so as to result unacceptable, especially for voice or audio services.
Another prior art document, the US patent No. 6,791,959 Bl in the name of Broadcom Corporation, provides a general-purpose method for improving the communication performance in a wireless communication comprising at least one mobile electronic device and a plurality of transmitter/ receiver sites. The method evaluates over time intervals the quality of the communication by acquiring complex measurements like signal-to-noise ratio, bit error rate, number of transmission retries, received signal strength indicator, bit energy-to- noise density, global position information and cell congestion information. Functional relationships based on successive measurement are compared and extrapolated in order to determine whether to roam, prior to a next transmission.
However, the parameters considered in such document are low OSI (Open Systems Interconnection) layer parameters, which are not indicated to monitor services provided by software applications at higher OSI layers, producing Constant Bit Rate traffic as for video streaming and audio service. Furthermore the evaluation of a numerous number of physical parameters, the storing of historical data and the extrapolated measurements required by this method, make it not adapt for low computational electronic mobile devices.
Also the article to Ishihara et al. entitled: "Predictive rate control for real time video streaming with network triggered handover" describes a control method which is based on the evaluation of the effectiveness of a predictive quality of service control scheme for real-time video streaming on a wireless network that supports network-triggered handover.
Even if such scheme may be used to monitor the wireless network channel when used by a software application, it relies on a wireless network provided by specific access points, able to perform some additional and high-layers functions; in other words the monitoring of a deterioration of the wireless network channel is demanded to additional devices and/ or additional functions of the access points and it is not entirely managed by the final-user electronic mobile device.
Another document, the article to Misikangas et al. entitled: "Predicting QoS for Nomadic Applications Using Intelligent Agents", describes a method which tries to detect in short-term the predictability of the available quality of service, in order to alert a user that a connection for the electronic mobile device is no more available.
However, also the method according to Misikangas requires very high resources to the electronic mobile device while it does not guarantee a response time for the prediction suitable to audio and video real time services.
The problem at the basis of the present invention is that of providing a method for detecting the deterioration of a wireless network channel, whereto an electronic device is connected through a wireless network interface and comprises a software application which produces a substantially Constant Bit rate traffic, useful for example in case of video or audio communications with one or more remote servers, the method providing predicting information for high OSI layer software applications before the communication becomes impossible, without requiring to the electronic mobile device computational intensive operations nor additional information received from said servers, nor from additional access points or devices belonging to the wireless network, overcoming all the problems that currently affect the detection of the deterioration in wireless channel according to the prior art.
Summary of the invention
The solution idea on which the present invention is based, is that of providing a method for predictive detection of a wireless network channel degradation, evaluating widely accessible parameters and functions related to a high OSI layer software application that is associated to a Constant Bit rate traffic, the method taking into account the average throughput transmitted /received by an electronic mobile device to/from the wireless network channel.
This problem is solved, according to the present invention, by a method for detecting a deterioration in a wireless network channel, whereto an electronic mobile device is connected through at least one wireless network interface, the electronic mobile device comprising a software application, for transmitting/ receiving a substantially Constant Bit Rate traffic to/from the wireless network channel, the method comprising
a signal strength sampling phase, for sampling an average strength of a signal, received by the at least one wireless network interface on the wireless network channel;
a detection phase, for detecting when the average strength reaches a predefined threshold value;
the method being characterized by comprising:
a bound sampling phase, defining a lower and an upper bound of a throughput threshold interval by sampling a throughput generated by the software application over the wireless network channel;
a throughput sampling phase, for sampling an average of the throughput,
the detecting phase comparing said average of the throughput with respect to said lower and upper bounds of the throughput threshold interval and issuing an alert signal to a final user thereby signaling said deterioration of the wireless network channel, before the lost of the communication .
A detection of the deterioration of the wireless network channel, whereto the electronic mobile device is connected, is determined before that the substantially Constant Bit Rate traffic produced by the software application is delayed to the user, and thus before rendering for example a video or audio communication unusable, such a detection not requiring computational intensive operations to the electronic mobile device nor additional resources external to it.
Further characteristics and the advantages of the method according to the present invention will be apparent from the following description of an embodiment thereof, made with reference to the annexed drawings, given for indicative and non-limiting purpose. Brief description of drawings
Figure 1: schematically shows, in a block diagram, the main phases of the method for predictive detection of deterioration in a wireless network channel according to the invention.
Figure 2: schematically shows, in a block sub-diagram 2, one phase of the method according to the invention.
Figure 3: schematically shows, in a block sub-diagram 3, another phase of the method according to the invention.
Figure 4: schematically shows, in a block sub-diagram 4, a further phase of the method according to the invention.
Figure 5: schematically shows, in a block sub-diagram 5, yet another phase of the method according to the invention.
Detailed description
According to the present invention and with reference to figure 1, a method for predictive detection of degradation in a wireless network channel, over which an electronic mobile device 1 communicates, is schematically indicated with block diagram 60.
In the following description, the term electronic mobile device 1 is referred to devices of the type comprising one or more physical wireless network interfaces 12, for supporting corresponding one or more network communications over wireless network channels.
For exemplificative purpose and without limiting the scope of the present invention, the electronic mobile device 1 could be a laptop, a PDA, a smart-phones or a similar devices.
Always for exemplificative purpose without limiting the scope of the invention, some example of physical wireless network interfaces 12 are interfaces for an optical network, like infrared networks, for Wi-Fi networks according to the IEEE 802.11 standard, for Bluetooth network according to the IEEE 802.15.1 standard, for ZigBee networks according to IEEE 802.15.4 standard, for GPRS networks, for EDGE networks, for IXRTT networks, for CDMA2000 networks, for UMTS networks, for HSDPA networks, for WiMAX networks according to the IEEE 802.16 standard and the like.
The electronic mobile device 1 comprises a software application 11 that, in execution, generates a substantially Constant Bit Rate traffic. For example an audio and/ or video application, having stringent time and bandwidth requirements, produces a continuous and almost Constant Bit Rate traffic of real time data. The software application 11 may also be a real time voice call application, a video call application, a video-on- demand application, a mobile TV applications, a mobile radio application or similar.
As represented with reference to block diagram 13 of figure 1, the electronic mobile device 1 also comprises a network communication module, for transmitting/ receiving a number of frames at Constant-Bit Rate traffic over the wireless communication channel, through the above mentioned interface 12.
According to the present invention, a detection phase 2 for detecting the wireless channel degradation may be activated on the electronic mobile device 1 in order to preventively determine the availability of the channel. More particularly, the detection phase may be recalled by a user, for example through a keyboard on the electronic mobile device 1, or by a system user, according to predefined conditions; the human user and the system user are schematically indicated in figures with numeral reference 7 and also indicated as end-consumer. The predictive detection can be used by an end-consumer 7 to bring into being all the necessary pro-active activities to minimize the quality of service reduction of the software application 11, i.e. the throughput reduction and the user experience worsening. Said pro-active activities require the application or its user intervention to react to the different and worse wireless channel conditions. Instead, if the end-consumer 7 is a seamless handover system the application or its user intervention can be avoided or minimized. A seamless handover system can, in fact, use the predictive detection of the wireless channel degradation to start the selection of a new suitable wireless channel, if present, and connect to it automatically, if possible, and transparently let the application continue to work as before minimizing the throughput reduction and the user experience worsening.
The detection phase 2, described later in more detail, is supported by the following phases:
- a signal strength sampling phase 3, intended to determine the average signal strength on the basis of samples of the signal strength retrieved from the at least one wireless network interface 12 with a predetermined time interval, hereinafter indicated as signal strength sampling time interval;
- a bound sampling phase 4, intended to determine an upper and a lower bound of a throughput threshold interval defined on the basis of the average of the throughput generated by the software application 11 over the wireless communication channel and sampled at a low sampling time interval;
- a throughput sampling phase 5, intended to determine the average of the throughput generated by the software application 11 over the wireless communication channel and sampled at a high sampling time interval;
a timing phase 6, triggering the execution of the above mentioned phases.
It should be remarked that the above described phases are to be considered as tasks or threads which are contemporary performed. More particularly, the detection phase 2, the signal strength sampling phase 3, the bound sampling phase 4 and the timing phase 6 are in execution till the signal strength value is above a warning threshold value, while the detection phase 2, the signal strength sampling phase 3, the throughput sampling phase 5 and the timing phase 6 are executed when the signal strength value overcomes such a warning threshold value, as will be more clearly described in the following lines.
Hereinafter a brief description of an embodiment of phases 3, 4, 5 will be described with reference to respective corresponding figures 3, 4, 5.
The signal strength sampling phase 3 is recalled by the detection phase 2 in order to calculate the average value of the signal strength on the basis of a plurality of values or samples stored in a first or signal strength buffer 31. The first time it is activated, the signal strength sampling phase 3 initializes the signal strength buffer 31 with a predetermined value, indicated as signal strength warning threshold value, for example stored in a non volatile memory portion of the electronic mobile device 1. The signal strength warning threshold value may be configurable, for example on the basis of statistics related to the specific wireless technology used for communication by the electronic mobile device 1.
The signal strength buffer 31 comprises a plurality of memory locations for storing a plurality of corresponding signal strength samples; more particularly, the signal strength buffer 31 is a rotational buffer, per se known in the art.
Without limiting the possibility of implementing the signal strength buffer 31 according to the invention and only for exemplificative purpose, it could be considered as a linear buffer wherein the samples of the signal strength are stored from left to right inside the signal buffer 31 itself. When the rightmost position is filled, a new sample may be stored in the leftmost position with the right shifting of all the previously stored data and the discarding of the rightmost one as it stores the oldest and less updated sample.
The signal strength sampling phase 3 stores data also in a second or average signal strength buffer 32, for example similar to the signal strength buffer 31 and in particular a rotational one. The first time it is activated, the signal strength sampling phase 3 initializes also the average signal strength buffer 32 with the signal strength warning threshold value.
After the initialization of the signal strength and the average signal strength buffers, 31 and 32, the signal strength sampling phase 3 waits for a trigger from the timing phase 6 in order to retrieve and evaluate new signal strength samples. When triggered, the signal strength sampling phase 3 retrieves from the at least one wireless network interface 12 at least a new sample and inserts it in the signal strength buffer 31. For example the first time, such sample may be stored in the leftmost position of the signal strength buffer 31 , the second time in the following position and so on.
On the base of the retrieved sample, the signal strength sampling phase 3 calculates an average value of the values i.e. the samples stored in the signal strength buffer 31 and stores such average value in the average signal strength buffer 32. Also for the average signal strength buffer 32, the first time, such average value may be stored in the leftmost position of the average signal strength buffer 32 itself, the second time in the following position and so on.
Such a signal strength sampling phase 3 performs the above described operations and samplings, each time it is triggered by the timing phase 6.
Now, with reference to figure 4, the bound sampling phase 4 will be described.
The bound sampling phase 4 is recalled by the detection phase 2 in order to calculate a throughput warning threshold interval, on the basis of a plurality of values stored in a third or bound buffer 41 and determined by the throughput associated to the software application 11. More particularly, the throughput warning threshold interval is calculated with boundaries that are fractions of the average of values stored in the bound buffer 41 , as described hereafter.
It should be noted that such a bound sampling phase 4 calculates the throughput warning threshold interval on the basis of the average throughput value when the software application 11 normally works (for instance, the throughput warning threshold interval currently used is between -9% and -15% of the average throughput value).
The first time it is activated, the bound sampling phase 4 initializes the bound buffer 41 with a predetermined value, for example "0" in all the locations or a value stored in a non volatile memory portion of the electronic mobile device 1; also this value may be configurable.
The bound buffer 41 comprises a plurality of memory locations for storing a plurality of corresponding values of the throughput; more particularly, the bound buffer 41 is a rotational buffer, for example of the type described with reference to the signal strength buffer 31. After the initialization of the bound buffer 41, the bound sampling phase 4 waits to be triggered by the timing phase 6; when triggered, it samples the throughput generated by the software application 11 and insert it in the bound buffer 41.
More particularly, when the bound sampling phase 4 is triggered by the timing phase 6, it may return in a waiting state if the detection phase 2 does not authorize it to proceed in the sampling of frames. This case, as it will be clear later, happens when the detection phase 2 considers the signal strength sufficiently low to require deeper analysis of the throughput.
The bound sampling phase 4 calculates boundaries of the throughput warning threshold interval on the basis of fractions of the average of values stored in the bound buffer 41. More particularly, the throughput warning threshold interval boundaries may be calculated according to different advantageous formulas and can also have the same value.
Such a bound sampling phase 4 performs the above described operations and sampling each time it is triggered by the timing phase 6.
Now, with reference to figure 5, the throughput sampling phase 5 will be described.
The throughput sampling phase 5 is recalled by the detection phase 2 in order to calculate, on the basis of a plurality of values stored in a fourth buffer 51, an average value of the throughput generated by the software application 11 as described hereafter.
The first time it is activated, the throughput sampling phase 5 initializes fourth and fifth buffers, 51 and 52, also indicated as throughput and average throughput buffers, with a predetermined value. Advantageously, such predefined value may be the upper bound of the throughput warning threshold interval calculated by the bound sampling phase 4.
The throughput and average throughput buffers 51 and 52 comprise a plurality of memory locations for storing a plurality of corresponding values. The throughput and average throughput buffers 51 and 52 are rotational buffers, as explained with reference to the signal strength buffer 31.
After the initialization of the throughput and average throughput buffers 51 and 52, the throughput sampling phase 5 waits to be triggered by the timing phase 6; when triggered, it samples the throughput generated by the software application 11 and insert it in the throughput buffer 51.
Also in this case, when the bound sampling phase 4 is triggered by the timing phase 6, it may return in a waiting state if the detection phase 2 does not authorize it to proceed in the sampling of frames. This happens when the detection phase 2 considers the signal strength sufficiently high to avoid deeper analysis of the throughput.
On the basis of the last throughput value sampled, the throughput sampling phase 5 calculates the average value of the values stored in the throughput buffer 51 and stores such average value in the average throughput buffer 52.
Such a throughput sampling phase 5 performs the above described operations and sampling each time it is triggered by the timing phase 6.
The operations performed by the signal strength sampling phase 3, the bound sampling phase 4 and the throughput sampling phase 5may be resumed as follows:
- the signal strength sampling phase 3 is responsible to sample signal strengths in order to evaluate the average strength of the signal respect to a signal strength threshold;
- the bound sampling phase 4 is responsible to sample at a low rate the throughput generated by the software application 11, in order to determine a throughput warning threshold interval, comprising a lower and an upper bound;
- the throughput sampling phase 5 is responsible to sample at a high rate the throughput generated by the software application 11, in order to determine an average value of said throughput, a starting value of said average value being related to the upper bound calculated in bound sampling phase 4, that corresponds to a potential starting point of a deterioration of the wireless network channel.
More particularly, the signal strength sampling phase 3 and the bound sampling phase 4, which are triggered by the timing phase 6, support the detection phase 2 in order to evaluate the degradation of the wireless network channel: when the detection phase 2 considers that the wireless network channel is working normally, as the signal strength is greater than the signal strength warning threshold value, only the signal strength sampling phase 3 and the bound sampling phase 4 are activated.
If the detection phase 2 does not detect any deterioration, the throughput sampling phase 5 does not work and the bound sampling phase 4 samples at a low rate the throughput in order to provide the throughput warning threshold interval as described: in this case, a warning counter is set to 0 in order to remember that the current state of the wireless network channel is of non deterioration.
When the detection phase 2 detects a deterioration (on the basis of the signal strength warning threshold value), it requires to the bound sampling phase 4 the last updated throughput warning threshold interval, stops the bound sampling phase 4 and activates the throughput sampling phase 5.
At this point, the detection phase 2 compares the lower and upper bounds of the last throughput warning threshold interval updated by the bound sampling phase 4 with the average throughput calculated by the throughput sampling phase 5.
More particularly:
- if the average throughput calculated by the throughput sampling phase 5 is lower than the lower bound of such interval, the detection phase 2 has detected a deterioration of the wireless network channel and issues an alert to the human or system user 7;
- if the average throughput calculated by the throughput sampling phase 5 is higher than the lower bound of the interval but it is lower than its upper bound and if the warning counter value is lower than a predefined or configurable critical value (for instance, a value equal to 1, as shown in figure 2), the detection phase 2 does not issue any alert to the human or system user 7 but increases the warning counter by 1, as an indication that the condition of the wireless network channel is probably worsening (i.e. it is probably deteriorating);
- if the average throughput calculated by the throughput sampling phase 5 is higher than the lower bound of the interval but it is lower than its upper bound and if the warning counter value is greater than or equal to the above mentioned critical value, the detection phase 2 has detected a deterioration of the wireless network channel and issues an alert to the human or system user 7;
if the average throughput calculated by the throughput sampling phase 5 is higher than the upper bound of the interval, the detection phase 2 set the warning counter to 0, as a feedback that the condition of the wireless network channel is returning in acceptable values (i.e. the wireless network channel is not deteriorated).
Hereinafter a more detailed description of the detection phase 2 and its interaction with phases 3, 4, 5 as above described will be given with reference to figure 2.
As stated, the detection phase 2, may be activated by a human user 7 in order to monitor the wireless network channel and to be alerted if a deterioration occurs. The detection phase 2 may also be activated by a system user 7, for example by a process requiring the monitoring of the deterioration of the wireless network channel (e.g. a seamless handover system) .
The timing phase 6 is set so as to activate an execution of the detection phase 2 at predefined events or intervals, for example:
- after a setup time or
- each time a specified execution time interval has elapsed.
The execution time interval, in order to avoid unnecessary executions of the detection phase 2 on unmodified data, may be greater than the minimum between the signal strength sampling time interval and the high sampling time interval.
In a preferred embodiment, in order to take advantage as soon as possible of the new signal strength and throughput sampled by the signal strength sampling phase 3 and the bound sampling phase 4, the execution time interval may be chosen so that the difference between its duration and the minimum between the signal strength sampling time interval and the high sampling time interval is substantially equal to the time required by the detection phase 2 to complete a single execution.
The timing phase 6 may be set by the signal strength sampling phase 3, the bound sampling phase 4 and by the throughput sampling phase 5.
More particularly:
- the signal strength sampling phase 3 sets the timing phase 6 so that a trigger may activate phase 3 itself in order to sample the signal strength on the wireless network channel; hereinafter such trigger is indicated as signal strength sampling trigger;
the bound sampling phase 4 sets the timing phase 6 with a low sampling trigger, so that a trigger may activate phase 4 itself in order to sample the throughput at low rate and to determine the corresponding lower and upper bounds of the throughput warning threshold interval;
- the throughput sampling phase 5 sets the timing phase 6 with a high sampling trigger, so that a trigger may activate phase 5 itself in order to sample the throughput at a high rate.
The bound sampling phase 4 is executed only when the wireless channel is in normal conditions, in other words when the throughput samples fluctuations are not relevant, while the throughput sampling phase 5 is used when the channel conditions are getting worst, when the short time determination of an average throughput negative trend may bring to a predictive detection of the wireless communication channel degradation.
Advantageously, also the high sampling time interval may be chosen so as to be greater than the time interval between two consecutive packets generated by the Constant Bit Rate traffic of the software application 11, in order to avoid a throughput sample equal to zero and so determining a false prediction of the wireless communication channel degradation.
The setup time in the timing phase 6, associated to the detection phase 2, is greater than or equal to the time requested by the signal strength sampling phase 3 and the bound sampling phase 4 to start in order to produce valid average values based on a reasonable number of real sampled data.
As schematically represented in the block diagrams of figure 2, after the setup time has elapsed and after the first execution trigger has been triggered from the timing phase 6, the detection phase 2 enter a first loop of execution recalling the signal strength sampling phase 3. The signal strength sampling phase 3 returns the last average signal strength calculated on the basis of the signal strength samples, provided by the at least one wireless network interface 12 of the electronic mobile device 1.
This average value is compared with a signal strength warning threshold value, that is for example a predetermined value or a configurable value, indicating the signal strength value at which it is required a more intensive analysis on the throughput generated by the software application 11. In other words, when the average value is higher than the signal strength warning threshold value it means that the signal is sufficiently high and no deeper analysis phases, in addiction to the signal strength sampling phase 3 and the bound sampling phase 4, are required in order to the detect the deterioration of the wireless network channel.
More particularly, above this signal strength threshold, the throughput generated by the software application 11 cannot be influenced by the signal strength variations because it is so strong that its strength fluctuations cannot have direct effects on the throughput.
On the contrary, at a critical value below this signal strength threshold, the throughput generated by the software application 11 could be heavily influenced by the signal strength variations, for example when the electronic mobile device 1 approaches a border of a wireless cell whereto it is currently connected.
In this case, advantageously according to the invention, an evaluation of physical layer parameters is only used as a trigger for a deeper analysis involving OSI data link and/or OSI network and/or OSI transport and/or OSI application layer parameters collected at a proper sampling rate.
More particularly, when the average signal strength falls down the signal strength warning threshold, a deeper analysis of the throughput as generated by the software application 11 is started. Such a deeper analysis is carried out by using values that are sampled at a higher rate than a normal condition rate, normally used to calculate the reference average throughput. This critical signal strength value is typical for each specific wireless technology since the signal strength has different power and different characteristics from one wireless technology to another. Furthermore, inside a same wireless technology, this critical signal strength value varies depending on the hardware and software of the wireless interfaces, for example depending on the specification followed by the different manufacturers or on the interface sensitivity.
The signal strength warning threshold, instead, depends only on the wireless technology considered and not on the particular wireless interface used and it can be derived from statistical studies for instance as a fraction of the average critical signal strength thresholds.
So, in normal condition, the last average signal strength is greater than or equal to the signal strength warning threshold value. In this normal condition, in order to avoid unnecessary execution loops of the detection phase 2 on unmodified data, and in particular when the execution time interval set in the timing phase 6 is lower than the signal strength sampling time interval and greater than the high sampling time interval, the timing phase 6 may be reset to trigger the detection phase 2 after a sleep time has elapsed. This sleep time should be greater or equal to the difference between the signal strength sampling time interval and execution time interval.
An abnormal condition, corresponding to a potential degradation of the wireless network channel, occurs when the last average signal strength is lower than the signal strength warning threshold value.
In this case, the detection phase 2 requests to the bound sampling phase 4 the last throughput warning threshold interval calculated. This interval has boundaries calculated as fractions of the average low sampling throughput derived from the number of frames generated by the constant-bit rate traffic in the low sampling time interval and provided by an operating system network communication module 13 of the electronic mobile device 1.
When the detection phase 2 has received the last throughput warning threshold interval, it starts the throughput sampling phase 5, if not already running, and requests to it the last average throughput calculated.
This value is derived from the number of frames generated by the constant-bit rate traffic during the high sampling time interval and provided by an operating system network communication module 13 of the electronic mobile device 1.
If the last average throughput is lower than the lower bound of the received throughput warning threshold interval or it is for at least two times consecutively lower than the upper bound but greater than the lower bound of the received throughput warning threshold interval, the detection phase 2 detects the degradation of the wireless network channel before the communication over it becomes impossible and signals it to the user 7, for example to an end-consumer.
In order to achieve the predictive detection of the wireless communication channel degradation the constant-bit rate throughput generated by the software application 11 must be much greater (at least one order of magnitude) than the eventual average variable-bit rate throughput generated by the mobile device operating system and/ or by the eventual other applications running on the electronic mobile device 1.
This eventual variable-bit rate traffic can be considered as noise and this noise has to be much lower than the constant signal represented by the constant-bit rate traffic generated by the software application 11 , in order to detect with reasonable sureness when the throughput starts to fall, before the communication becomes impossible.
The present invention also relates to an electronic mobile device 1 for predictive detection of deterioration in a wireless network channel.
In order to simplify the description of such an electronic mobile device 1, the hardware components of the same, intended to implement the phases of the method above described, will be indicate by the same referral numbers. The electronic mobile device 1 thus comprises:
- an operating system network communication module 13, for transmitting/ receiving the frames of the Constant-Bit Rate traffic to /from the wireless network channel,
at least one wireless network interface 12, for connecting the electronic device 1 to the wireless network channel,
a pre-stored software application 11, generating a substantially constant bit rate traffic,
detecting means 2 for alerting a user 7 of the application about said deterioration.
More particularly, according to a preferred embodiment of the present invention, the electronic mobile device 1 also comprises the following:
signal strength sampling means 3 for sampling an average strength of a signal, received by the at least one wireless network interface 12 on the wireless network channel;
bound sampling means 4 defining a lower and an upper bound of a throughput threshold interval by sampling a throughput generated by the software application 11 ;
throughput sampling means 5 for sampling an average of the throughput.
The detecting means 2, when the average strength of the signal reaches a predefined threshold, raises the predictive detection if the average throughput is out of bound with respect to the lower and upper bounds of the throughput threshold interval.
The electronic mobile device 1 also comprises a timer 6 for triggering, at corresponding predefined intervals, executions of
the detecting means 2,
- the signal strength sampling means 3,
the bound sampling means 4,
the throughput sampling means 5.
The detecting means 2, signal strength sampling means 3, bound sampling means 4 and throughput sampling means 5 set the corresponding predefined intervals in the timer 6.
More particularly, the signal strength means 3 samples a plurality of values of the signal strengths and stores it in a first or signal strength buffer 31, comprising a corresponding plurality of memory locations; averages of the plurality of values are stored in successive memory locations of a second or average signal strength buffer 32.
The bound sampling means 4 samples a plurality of throughput values and stores it in a third or bound buffer 41 , comprising a corresponding plurality of memory locations, the lower and upper bounds of the throughput threshold interval being determined on a computation of the plurality of the throughput values.
The throughput sampling means 5 samples a plurality of throughput values and stores it in a fourth or throughput buffer 51, comprising a corresponding plurality of memory locations; averages of the plurality of the throughput values being stored in successive memory locations of a fifth or average throughput buffer 52.
According to a preferred embodiment of the electronic device, the detecting means 2, the signal strength sampling means 3, the bound sampling means 4 and the throughput sampling means 5 are implemented through corresponding hardware modules. Alternatively, these means can be part of a more complex apparatus or implemented by software applications. Advantageously, according to the method and to the device of the present invention, a detection of the deterioration of a wireless network channel, whereto an electronic mobile device is connected through a wireless network interface, producing a substantially Constant Bit Rate traffic, is determined in time to alert a user before the connection falls.
The detection according to the present invention is particularly advantageous in video or audio communications between the electronic mobile device and one or more remote servers, especially in predicting deterioration in a wireless network channel used by high layer software applications before the communication becomes impossible.
The method and device does not require computational intensive operations or additional information to said remote server, or to additional access points or devices belonging to the wireless network, and it is based on the resource normally available in electronic mobile devices.

Claims

1. Method for detecting a deterioration in a wireless network channel, whereto an electronic mobile device (1) is connected through at least one wireless network interface (12), said electronic mobile device (1) comprising a software application (11), for transmitting/ receiving a substantially constant bit rate traffic to/from the wireless network channel, the method comprising
- a signal strength sampling phase (3), for sampling an average strength of a signal, received by said at least one wireless network interface (12) on said wireless network channel;
a detection phase (2), for detecting when said average strength reaches a predefined threshold value;
the method being characterized by comprising:
- a bound sampling phase (4), for defining a lower and an upper bound of a throughput threshold interval by sampling a throughput generated by said software application (11) over said wireless network channel;
- a throughput sampling phase (5), for sampling an average of said throughput,
said detecting phase (2) comparing said average of the throughput with respect to said lower and upper bounds of the throughput threshold interval and issuing an alert signal to a final user (7) thereby signaling said deterioration of the wireless network channel, before the lost of the communication .
2. Method according to the claim 1 characterized by comprising a timing phase (6) for triggering executions of said detection phase (2), said signal strength sampling phase (3), said bound sampling phase (4) and said throughput sampling phase (5), at corresponding predefined intervals.
3. Method according to claim 2 characterized by the fact that said signal strength sampling phase (3) samples a plurality of values of said signal strengths and stores it in a signal strength buffer (31), comprising a corresponding plurality of memory locations, averages of said plurality of values being stored in successive memory locations of an average signal strength buffer (32).
4. Method according to claim 3 characterized by the fact that said average strength of the signal is calculate by an average of said averages of said plurality of values stored in said average signal strength buffer (32).
5. Method according to claim 4 characterized by the fact that said bound sampling phase (4) samples a plurality of throughput values and stores it in a bound buffer (41), comprising a corresponding plurality of memory locations, said lower and upper bounds of the throughput threshold interval being determined on a computation of said plurality of throughput values.
6. Method according to claim 5 characterized by the fact that said throughput sampling phase (5) samples a plurality of throughput values and stores it in a throughput buffer (51), comprising a corresponding plurality of memory locations, averages of said plurality of throughput values being stored in successive memory locations of an average throughput buffer (52).
7. Method according to claim 6 characterized by the fact that said average of said throughput is calculated by an average of said averages stored in said average throughput buffer (52).
8. Method according to the claim 7 characterized by the fact that said bound sampling phase (4) and said throughput sampling phase (5) retrieve the throughput samples of the traffic generated by the software application (11) from an operating system network communication module (13) of the electronic mobile device (1).
9. Electronic mobile device (1) for predictive detection of a deterioration in a wireless network channel, the electronic mobile device (1) comprising at least one wireless network interface (12), for connecting to the wireless network channel, a pre-stored software application (11), generating a substantially constant bit rate traffic, an operating system network communication module (13) for transmitting/ receiving said constant bit rate traffic to/ from said wireless network channel, detecting means (2) for alerting a final user (7) of the application about said deterioration,
characterized by comprising:
signal strength sampling means (3), for sampling an average strength of a signal, received by said at least one wireless network interface (12) on said wireless network channel;
bound sampling means (4), for defining a lower and an upper bound of a throughput threshold interval by sampling a throughput generated by said software application (11);
throughput sampling means (5), for sampling an average of said throughput,
said detecting means (2) interfacing said signal strength sampling means (3), said bound sampling means (4), said throughput sampling means (5) and issuing an alert signal to the final user (7) thereby signaling said deterioration of the wireless network channel, before the lost of the communication.
10 Electronic mobile device (1) according to claim 9 characterized by comprising a timer (6) for triggering executions of said detecting means (2), said signal strength sampling means (3), said bound sampling means (4), said throughput sampling means (5), at corresponding predefined intervals.
11. Electronic mobile device (1) according to claim 10 characterized by the fact that said signal strength sampling means (3) samples a plurality of values of signal strengths and stores it in a signal strength buffer (31), comprising a corresponding plurality of memory locations, averages of said plurality of values being stored in successive memory locations of an average signal strength buffer (32).
12. Electronic mobile device (1) according to claim 11 characterized by the fact that said bound sampling means (4) samples a plurality of throughput values and stores it in a bound buffer (41), comprising a corresponding plurality of memory locations, said lower and upper bounds of the throughput threshold interval being determined on a computation of said plurality of throughput values.
13. Electronic mobile device (1) according to claim 12 characterized by the fact that said throughput sampling means (5) samples a plurality of throughput values and stores it in a throughput buffer (51), comprising a corresponding plurality of memory locations, averages of said plurality of throughput values being stored in successive memory locations of an average throughput buffer (52).
PCT/EP2006/011548 2005-12-01 2006-12-01 A method for predictive detection of deterioration in a wireless network channel WO2007062859A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
EP05026479 2005-12-01
EP05026479.5 2005-12-01

Publications (1)

Publication Number Publication Date
WO2007062859A1 true WO2007062859A1 (en) 2007-06-07

Family

ID=37891956

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/EP2006/011548 WO2007062859A1 (en) 2005-12-01 2006-12-01 A method for predictive detection of deterioration in a wireless network channel

Country Status (1)

Country Link
WO (1) WO2007062859A1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009015176A1 (en) * 2007-07-23 2009-01-29 Telcordia Technologies, Inc. Method and system for improving wireless customer experience by anticipating and explaining communication quality problems through notifications
CN114362838A (en) * 2020-09-28 2022-04-15 安立股份有限公司 Mobile terminal testing device and mobile terminal testing method

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030033394A1 (en) * 2001-03-21 2003-02-13 Stine John A. Access and routing protocol for ad hoc network using synchronous collision resolution and node state dissemination
US6556553B1 (en) * 1999-04-12 2003-04-29 Intermec Ip Corp. Method for determining when a communication device should rate shift or roam in a wireless environment
US20050169186A1 (en) * 2004-01-30 2005-08-04 Microsoft Corporation What-if analysis for network diagnostics

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6556553B1 (en) * 1999-04-12 2003-04-29 Intermec Ip Corp. Method for determining when a communication device should rate shift or roam in a wireless environment
US20030033394A1 (en) * 2001-03-21 2003-02-13 Stine John A. Access and routing protocol for ad hoc network using synchronous collision resolution and node state dissemination
US20050169186A1 (en) * 2004-01-30 2005-08-04 Microsoft Corporation What-if analysis for network diagnostics

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009015176A1 (en) * 2007-07-23 2009-01-29 Telcordia Technologies, Inc. Method and system for improving wireless customer experience by anticipating and explaining communication quality problems through notifications
KR101285741B1 (en) 2007-07-23 2013-07-19 텔코디아 테크놀로지스, 인코포레이티드 Method and system for improving wireless customer experience by anticipating and explaining communication quality problems through notifications
CN114362838A (en) * 2020-09-28 2022-04-15 安立股份有限公司 Mobile terminal testing device and mobile terminal testing method

Similar Documents

Publication Publication Date Title
JP5197363B2 (en) Method and apparatus for providing high quality services between terminals based on performance characteristics of network conditions
KR101107945B1 (en) Reducing packet loss for a packet data service during congestion in a transport network
US9125138B2 (en) System and method for optimizing video conferencing in a wireless device
US8516101B2 (en) Resource management for a wireless device
US7508760B2 (en) Data distribution server and terminal apparatus
CN101160848B (en) Method and device for dynamic control data clustering transmission
US7218929B2 (en) Method and system for improved monitoring, measurement and analysis of communication networks utilizing dynamically and remotely configurable probes
CN100471140C (en) Method for detecting QoS
EP1604475B1 (en) Method for obtaining information about a transmission capability
EP3841787B1 (en) Methods and apparatus for proactively switching between available networks
KR20080074793A (en) Media server and process of streming content
CN107277568A (en) A kind of plug-flow configuration parameter dynamic adjusting method and device
US20090135733A1 (en) Communication device, communication system, and method for monitoring communication quality between communication devices
WO2007062859A1 (en) A method for predictive detection of deterioration in a wireless network channel
CN105162719B (en) The packet loss detection system and method for Packet Transport Network
CA2236407C (en) Method and apparatus for performing automatic synchronization failure detection in an atm network
US20120083220A1 (en) Determining the quality of service in a communication terminal device
US20220264360A1 (en) Techniques for adaptive bitrate video traffic shaping
JP4065153B2 (en) Packet transmission apparatus and transmission timing control method
EP0974231B1 (en) Link adaption related to radio
CN1985488B (en) Data unit receiver, sender and method for controlling them
KR100676564B1 (en) Communication apparatus
JPH10248090A (en) Speed detection system utilizing fading in mobile communication
RU2364025C2 (en) Method and device for virtual unidirectional channel
US20220264375A1 (en) Network entities for enhanced qos monitoring

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application
NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 06818947

Country of ref document: EP

Kind code of ref document: A1