WO2008105684A1

WO2008105684A1 - A wimax network incorporating a mimo network to network technique

Info

Publication number: WO2008105684A1
Application number: PCT/SD2007/000004
Authority: WO
Inventors: Geili Tawfieg Abdalla El Sanousi
Original assignee: El Sanousi Geili Tawfieg Abdal
Priority date: 2007-02-28
Filing date: 2007-09-22
Publication date: 2008-09-04

Abstract

This invention relates to a method to implement MIMO in wireless Network to Network communications. The method is incorporated in a WIMAX architecture intended for macrocell broadband services. MIMO is realized through virtual multipath created by large scale integration of network RF terminals as virtual antenna units. The WIMAX architecture is organized into three layers: Base stations, Access points and End users. The method is deployed in communication between these layers (End users/ Access points in a MIMO-MU model and Access points/ Base stations in a MIMO-SU model). The Access points' layer therefore acts as interface between the two layers. The architecture deploys SFN-MIMO-OFDM baseband frame and uses fast link feedback between MIMO processors to enhance TDD.

Description

FIELD OF INVENTION

[001] This invention relates generally to a method to implement MIMO in outdoor

network to network communications through creation of artificial multipath (Non Line Of Sight-

NLOS) environment and large scale integration of the network tranceiving units (access points) as virtual antennas; more particularly it relates to a practical implementation of this method which is a

WIMAX network architecture intended for urban scale of users.

PRIOR ART AND THE MODIFICATIONS/DEVELOPMENTS TO PRIOR ART

[002] In mid 90(s) research lead to establishing the space time premise and introduced a new degree of freedom to the C. Shannon capacity equation which is the space dimension. It was as well rightfully established that the best exploitation of this dimension is the use of multi-element antennas both in transmission and reception (Multi-Input-Multi-Output-MIMO). MIMO so far has wider indoor WLAN applications than MIMO in outdoor; this is because outdoor applications are subject to many complexities such as multi user complexities, far-near problem, power constraints and so on.

[003] In the year 2005 Y. Takatori et al in their "Channel Capacity of TDD-OFDM-MIMO for

Multiple Access Points in a Wireless Single-Frequency-Network", Wireless Personal

Communications, VOL 35. pp 19-33, 2005, introduced a method to integrate WLAN access points to act as a large scale integrated virtual beamforming antenna array. As shown in this disclosure this concept is readily applicable to outdoor networks' architectures, particularly for increasing the capacity of the access spots (e.g. base stations, VSAT terminals and Satellite ground stations) rather than only end users

[004] The prior method combined the spatial multiplexing with Orthogonal Frequency Division

Multiplexing (OFDM) for broadband performance (MIMO OFDM) with a Single Frequency Network

(SFN) Baseband frame and Time Division Duplexing (TDD) with the perfect auto calibration method (ACT) described by T. K .Y Lo, "Maximum Transmission ratio ", Communications, Oct 1999. The artificial NLOS was produced by use of

Multiple MAP(s) all linked to an Access Control Unit (ACU) for MIMO and modulation algorithms (SVD or Schmidt gram as later proposed by same authors as improvement to the method). [005]The method was described and limited to indoor WLAN in a point to multipoint MIMO-MU (Multi User) mode between access points and mobile terminals (users). This disclosure exploits the potential of the method in a point to point single user (SU) MIMO for the purpose of communication between networks from different sources or those parts of the network that distribute to other parts of the same network.

[006]It is also established in literature that there is an upper limit beyond which the increase in number of transmission receive pairs does not increase the capacity of the channel (I. E. Telatar, "Capacity of multi-antenna Gaussian channels," Euro. Trans. Tel., vol. 10, no. 6, pp. 585-595, Nov./Dec. 1999. and Siriam Vishwanath, "Multi-User Multiple Antenna Systems: Theoretical limits and practical strategies", PhD thesis, Dept of Electrical and Computer Engineering, Brigham Young University, 2004, unpublished). The prior art did not treat this issue nor provided practical solutions to it. In this disclosure the channel matrix is limited to a specific number of operating terminals (virtual antenna elements) per transmission/reception. This can be achieved through antenna selection or beam selection, both being used in this invention.

[007]Perfect calibration and transmitter channel knowledge in outdoor application will be difficult to achieve because of the length of the time and angular spread of the channel. The present disclosure links the access control units of the different networks (layers) with high speed optical fiber connection for the purpose of transmitting channel state information and calibration weights for the ACT function (see [004]). Alternatively since terminals are fixed in position a look up table of channel state numbers and auto calibration weights for specific angles of

signatures can be developed (through campaigns) and stored in the processing units.

SUMMARY

[008]A WIMAX architecture with a promise of huge capacity, high data rates and enhanced performance is invented. The architecture is based on the concept of creating artificial NLOS in network to network communications. It uses MIMO-OFDM to provide broadband performance and is comprised of two communication subsystems. One subsystem incorporates single frequency network

(SFN) frame; thus provide for economy in frequency spectrum. The other subsystem also incorporates

SFN frame but may alternatively be implemented with sectored architecture. The sectored architecture is basically designed for WI-FI and is licensed to XIRRUS. To the knowledge of the inventor, at the time of this invention, the architecture is undergoing research by Motorola Inc for WIMAX outdoor applications. . The selection of WIMAX 802.16d provide a scalable volume for incorporation of

MIMO-OFDM with large FFT capacities and a room for processing the associated algorithms

(Schmidt-Gram or Singular value decomposition-SVD). TDD and perfect calibration are improved by the fast optical fiber connection for feedback. The network is designed based on pure geographical allocations rather than cellular technology. It is expected to provide huge capacities and high quality performance at both the access terminals and the end terminals.

BRIEF DESCRIBTION OF THE DRAWINGS

[009]Figure 1 is a graphical representation of the network signaling hierarchy, showing the three constituent layers of the network hierarchy (BSL, APL and EUL).

[OlOJFigure 2 is a schematic representation of geographic outlay of the BSL constituents

[Ol lJFigure 3 is a schematic representation of geographic outlay of the APL constituents.

[012]Figure 4 is a schematic representation of the overall network outlay. [013]Figure 5 illustrates the dual mode of operation of the APL when architecture.

[014] Figure 6 is a schematic representation of the BSL and APL larges scale system integration. The base stations are given the acronym Bi and access points are given the acronym Ai (i=l,2,3,..), both are shown as antennas.

[015]Figure 7 is a schematic representation of the APL- EUL communication model using MIMO-

MU. The users are given the acronym Si (i=l,2,3,..) and possess MultiElementAntenna (MEA) in their design.

[016]Figure 8 is a schematic representation of Satellite beams in a point to point MIMO link with

Satellite Earth stations.

[017]Figure 9 is a schematic representation of VSAT(s) in a point to point MISO link to a satellite (it can be MIMO if the satellite incorporated MEA).

[018]Figure 10 is a schematic representation of the BGAN units' cooperation in a transmit diversity to the satellite.

DETAILED DESCRIBTION OF THE INVENTION

[019]The present disclosure provides an improved wireless communication network benefiting from the spatial dimension introduced in the MIMO Space-Time premise. The disclosure describes a

WIMAX 802.16d network architecture although it is understood the approach can be applied to various outdoor networks and would easily lend itself to current trends towards integrating the existing networks into a global unified single network with different operators and MPLS.

[020]The network comprises of three categories of end terminals grouped into three layers; Base

Stations Layer (BSL), Access Points Layer (APL) and End User Layer (EUL). These three layers communicate in hierarchical manner [see fig 1] whereby in the downlink the BSL transmits to the APL and APL distributes to EUL. On the distributes back to BSL. The APL therefore

communication between BSL and APL uses two single RF carrier frequencies in the up and down links (with baseband SFN frames) thus provide for economy in spectrum. The APL-EUL communication has two alternate applications; when using MIMO-MU with TDMA model the network is also single frequency in the uplink and in the downlink. When using sectored architecture the SFN frame is not applicable and each sector will operate at a different frequency to the neighboring sectors.

[021] The BS layer is comprised of a few of number of Base Stations allocated at high altitude points and in the outer circumference of the targeted zone as shown in Fig 2. Here a Base Station refers to a MIMO-OFDM transceiver unit comprised of a large number of antenna elements M and operating as a very high capacity unit utilizing the highest 802.16d transmission range (31 miles). Each BS has M MIMO-OFDM stream. Its basic functions are beamforming and RF interfacing. The BS(s) are linked to an Access Control Unit (ACU) through a high speed optical fiber connection and cooperate together in transmission/reception and beamforming/beam selection. The high altitude is because WIMAX uses microwave spectrum and hence require LOS. While the integrated system exploits artificial NLOS, LOS is maintained between the RF interface units (virtual antennas BS(s) and AP(s)). The channel matrix rank is forced to be constant by antenna selection switching of BS(s) and AP(s) and beam selection with Butler matrix within the BS(s) and AP(s) units. For example if a base station is comprised of 32 elements and is chosen to link access points of 8 elements each then the number of AP(s) linked to the base station is limited to 32/8 =4 in order to make the channel rank non deficient. Also the number of beams from different BS(s) focused on an AP is 8 and number of beams from different AP(s) focused at a BS is 32. [022]In paragraph [021] above the ACU does the function of a MlMO-SlJ transceiver

of data received/transmitted from the BS(s) via the high speed link. Such jobs include MIMO-OFDM encoding/decoding, MIMO modulation algorithm (SVD or Schmidt and Gram), Channel space time training, channel estimation, beamforming algorithm and TDD calibration and synchronization. The ACU controls the antenna selection and can be used in collision avoidance scheme by forcing specific antenna/beam selections. The ACU has a feedback fast speed optical link with the equivalent Junior ACU(JACU) at the APL for the purpose of fast transmission of CSI (or channel state numbers) and calibration weights. The ACU is as well the gateway of the network to other networks (e.g. PSTN). [023] The APL is comprised of a greater number of AP(s) allocated at points of altitude that maintain the LOS with the BS(s). The allocations are dispersed in the targeted region -Fig 3- (this may follow a symmetrical geometry). An AP refers to a MIMO-OFDM transceiver unit comprised of a smaller number of antenna elements N and N MIMO-OFDM streams. It operates as average capacity unit utilizing the 802.16d 8-10 miles transmission range. More than one AP at a time communicates with a BS, the number is such that the channel matrix in not rank deficient. The method for partitioning arrays into virtual array [PCT/US/2006/022315] is used both at BS(s) and AP(s) to benefit from diversity gain, array gain and eliminate rank deficiency in virtual arrays at the BS-AP link where LOS is otherwise prominent. The AP layer acts as a switch between BSL and EUL, therefore it has dual communication algorithms. The frequency set (The uplink and Downlink) used between AP(s)-BS(s) is different from the set between AP(S)-EU(s). This prevents ambiguity in spatial signatures as a result of multipath and long distances. This also provides isolation between the three layers during operation. The AP(s) are linked via high speed optical fiber to the JACU which performs similar tasks to the ACU. The JACU also does the task of switching information streams between BSL and EUL (see Fig 4). [024]In paragraph [023] above the AP-EU communication is achieved by a whereby the AP(s) cooperate in transmission/reception with The JACU

processing and dynamic beamforming and a TDMA multiple access scheme. Alternative to this, is implementing sectored architecture at AP-EU while maintaining the AP-BS communication as MIMO-SU (Fig 5). This poses a question on support to mobility, frequency economy and associated handover problems.

[025] The EU(s) are assumed fixed although it is understood that with slight modification this layer could comprise of both mobile and fixed terminals MATHEMATICAL MOELS

[026]The mathematical model for the BSL-APL is derived from the general MIMO-OFDM model for a blind channel in "H. Boelcskei, D. Gesbert, and A. Paulraj, "On the capacity of wireless systems employing OFDM-based spatial multiplexing " IEEE Trans. Commun., vol. 50, pp. 225—234, Feb. 2002" developed to account for transmitter knowledge of channel (via the feedback link). The channel between the BSL and APL mimics a perfect virtual P2P MIMO system (Fig 6). For each tone there is a corresponding M_τ x M_R Channel Impulse Response (CIR). The capacity for the channel when perfectly known at transmitter is given as (for an independent single cluster realization):

Where ^ ⁼ ^^aS\^_k H=_O are the covariance matrices of the input vector (power), H is channel matrix (NM_RXNMT), M is the number of antenna elements and R,T denote receiver and transmitter respectively, N is the number of OFDM tones, I is identity matrix and P is total constrained power.

Using the identity det (I_mn+ AB) =det (I_nm + BA) on the above:

And since Σ is diagonal the expression inside reduces to

And building on the above reference model for blind channel we get

assigned to optimize power using water-pouring algorithm. Here where Hw is drawn from an

m variable Ν (0, 1), L is the number of clusters and R/ is the /^th cluster correlation matrix given by:

I And the 'correlation function':

_A[5Δ,^ή)]=e-^^Δcos*e^4(2OTΔsin(^^⁾²

s Δ = (n-m) Δ is the spacing in wavelengths between an array elements and θ, is the mean arrival

angle drawn from a Gaussian normal Ν (0, G_Q ).

[027]The link between the APL and EUL represents a MAC/BC MIMO-MU model (Fig 7). The capacities are defined by sum capacity and capacity region. However, there is a guaranteed linear growth in this sum capacity with increase in number of antenna elements at each terminal (AP(s) or EU elements); hence accommodates more users. The capacity region is a function of u [.]. Equations to estimate the capacity region and sum capacity

Vishwanath in [006]]. For this stage of the invention, the mathematical model of the original method of paragraph [003] will be satisfactory.

[028]The link between a BS and an AP maintains LOS. Although both have MEA but the LOS will undermine MIMO multiplexing gain between them (not the large scale integration MIMO). The BS(s) and AP(s) will achieve diversity and array gain by implementing the "method and system for partitioning an antenna array and applying Multi-Input Multi-Output and beamforming mechanism",

PCT/US/2006/022315 of NAVINI Networks, INC, of Jin, Hang and Grabner, John.

[029]Although the above disclosure is an application of a method for creating artificial NLOS in a

WIMAX urban network; the method can be applied to a variety of networks. Examples for this are enlisted below.

TERRESTRIAL SATELLITE NETWORKS

[030] Satellite networks communicate to earth in cellular like beams. These beams cover wide geographic areas and are characterised by strong LOS since no scatterers in space. The main loss is attributed to free space loss defined by Friis transmission equation and other attenuation causes. Using receiving terminals (ground stations) at geographically separated beams (implying NLOS) to exploit artificial diversity in a manner similar to in HF Australian Long-Fish Synchronous Multi-Channel

Access Signaling (SMASH) network architectures and the Modernized HF Communication Systems

(MHFCS) will boost the performance of such networks. However there is an associated delay caused by control signals transmission and processing flow. Effective Time Division Duplexing and channels state knowledge are trivial in applying the concept to this context; it is practical to assume a blind channel. TDD will be needed for synchronisation only since channel knowledge is only influential for the terminal (antenna) to beam provide power economy at the

reception. Fig 8 sketches the basic concept applied. VERY SMALL APERTURE TERMINALS (VSAT(S))

[031] With Very Small Aperture Terminals (VSAT(s)) this technique can be implemented on the condition that the satellite has MEA and is capable of beamforming function. While the number of terminals increases, the allocation of beams becomes bulky. Since not all terminals communicate simultaneously the size of the channel matrix can be fixed and antenna/beam selection technique may be used dynamically. Yet precise channel knowledge will not be needed but rather the beaming algorithm may rely on two 'pointers'; one reference ID of the terminal and its geographical coordinates and the other is a channel state number. At the ground since the VSAT(s) are basically used to eliminate long distance links, the feedback to the processing unit (within Hub) could be wireless. Single frequency networking with OFDM or Direct Spread DS-MIMO in combination with CDMA may be used. DS-MIMO can be used for diversity gain since practical outdoor losses and delays undermine the achievable capacity gain. Another practical advantage of DS-MIMO-CDMA is feasibility of upgrading existing networks since satellites are expensive to upgrade modulation units. The model is generally MIMO-MU where the mobile terminal (point) is the Satellite unit. Fig 9 illustrates the model. If the satellite unit comprises of MEA the model may then become point to point MIMO-SU.

BROADBAND GLOBAL AREANETWORKS

[032]Broadband Global Area Network (BGAN), the technology owned by INMARSAT constellation group has direct LOS communication with satellite and a support repeater earth station which also performs some similar functions to hub in VSAT networks. Exploiting multiplexing gain is not useful

Therefore exploiting diversity gain (from the user terminals transmission) will be the better alternative. This will turn the network independent of terrestrial ground station since the control unit will only coordinate the retransmission of data to the satellite. (At present local terrestrial ground stations are used as a link to strengthen user uplink signal). From practical point of view system adaptation to this is likely feasible since recently sent satellite units would have adaptive processing algorithms incorporated. Fig 10 below illustrates the model.

[033]The above disclosure provides many different embodiments or examples for implementing the artificial NLOS in networks' architectures. Also a specific application (A WIMAX network) with components and processes are described to explain the invention. These are examples and an application which are not intended to limit the invention from those described in the claims. It will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from spirit and scope of the invention. It is also fully understood, with respect to the WIMAX application that it incorporated a method for partitioning an antenna array and proposed a sectored architecture which are patents of referenced owners who are entitled to their protected patent rights upon implementation.

Claims

WHAT IS CLAIMED IS:

1. A method for wireless outdoor and macrocell network to network communications and wireless communication between network terminals (Network wireless distribution within a single hierarchical network). The method is an augmentation of [Y. Takatori et al, "Channel Capacity of

TDD-OFDM-MIMO for Multiple Access Points in a Wireless Single-Frequency-Network", Wireless Personal Communications, VOL 35. pp 19-33, 2005] in outdoor/macrocell nέtwδFtehrε: context; the method comprising:

Using the network RF interface units (access spots) as virtual antenna elements whereby each unit represents an antenna in a large scale integration of the network into a single system. The large scale integration transforms the networks into two (or more) network to network system or a wireless communication system between "same network'V'different hierarchy terminals" (wireless distribution within a single hierarchical network).

Virtual antenna arrays in effect produce a virtual Non Line Of Sight (NLOS) environment and hence generally provide for implementing Space Time (ST) mechanisms; particularly MIMO mechanisms. The array elements cooperate in beamforming, transmission and reception. Arrays of the integrated systems are linked via fast link to ST/MIMO processing units (Access Control Units) at the transmitter and at the receiver sides in the integrated systems respectively.

The number of virtual transmit/receive antenna elements pairs in MIMO channel is predetermined and made fixed through use of antenna selection switch, the channel matrix is hence not rank deficient indifferent of the number of terminals in the network.

A High speed optical fiber link is used as a feedback line between virtual transmitter and virtual receiver for the purposes of perfect auto calibration, synchronization and TDD channel state information to provide perfect knowledge of channel state at transmitter.

2. The method of claim 1 wherein the system MIMO mechanism is spatial multiplex scheme.

3. The method of claim 1 wherein the system MIMO mechanism is diversity transmit scheme.

4. The method of claim 1 wherein the system MIMO mechanism is beamforming for array gain.

5. The method of claim 1 wherein the system ST mechanism is optimum combining ratio with maximum-minimum technique.

6. The method of claim 1 wherein the system ST mechanism is spatial transmitter at a time.

7. The method of claim 1 wherein when the positions of the virtual antenna transmit/receive pairs are fixed, the feedback line is replaced by look up tables of channel state numbers and auto calibration weights for specific angles of arrival/departures and spatial signatures. These tables are developed through campaigns and afterwards stored in the MIMO processing units.

8. The method of claim 1 wherein the positions of the virtual antenna transmit/receive pairs experience strong LOS such that MIMO between the individual pairs is undermined; the method of partitioning arrays into virtual antenna arrays to create "local" artificial diversity [PCT/US/2006/022315] is used.

9. A WIMAX network architecture for use in urban and macro cells based on method of claim 1, The architecture comprising:

Three layers of similar classes of terminals grouped in each layer. The Base Stations and an Access control unit form the Base Station Layer (BSL). The Access Points and the Junior ACU (JACU) form the Access Points Layer (APL). The End User terminals form the End User Layer (EUL).

10. The architecture of claim 9 wherein the BSL is geographically the outer most layer at the circumference of the targeted zone, and acts as an interface between the network and external networks via its ACU. The BS(s) are positioned at high altitudes such as to maintain LOS with corresponding AP(s).

11. The architecture of claim 9 wherein the APL is geographically located at dispersed spots within the targeted zone and acts as an interface between End Users Level and the BSL. The altitude of these units is such that LOS is maintained with corresponding Base Stations.

forms of transmission and reception.

13. The architecture of claim 9 wherein the EUL is also geographically located at dispersed spots within the targeted zone and terminals are mobile. The terminals are independent of each other in all forms of transmission and reception.

14. The architecture of claim 9 whereat the interface between BSL and APL; the two layers are integrated into a point to point MIMO-SU system model. The interface uses single RF carrier frequency in the uplink and another different frequency in the downlink while the baseband frames at the two frequencies are of Single Frequency Network form. The total number of RF frequencies used in the interface is only two.

15. The architecture of claim 9 at the interface between APL and EUL; the two layers communicate in a point-to-multipoint, MIMO-MU system model. The interface also uses single RF carrier frequency in the uplink and another different frequency in the downlink (different from those in claim 11) while the baseband frames at the two frequencies are of Single Frequency Network form. The total number of RF frequencies used in the interface is also two. The multiple access scheme is TDMA.

16. The architecture of claim 9 at the interface between APL and EUL; the two layers can alternatively communicate in the sectored architecture of the reference sited in [008] and the model therefore is not a MIMO therefore each sector will have its frequency with no SFN baseband frame.

17. The architecture of claim 9 wherein the BS units are large capacity access spots, MIMO- OFDM transceivers with a Multi Element Antenna sets each of M elements. The M elements substantially outnumber the number of elements N comprised in the AP unit. modems corresponding to these elements and M streams are

passed to the processing from each BS. The BS basic functions include but not limited to: acting as a virtual RF interface (antenna), the 'local' beam switching and 'local' virtual array partitioning, and beamforming with other elements of the 'global' virtual array (of BS(s)). The word 'local' refers to the BS M element MEA and 'global' refers to BS(s) as part of the integrated system MEA. The BS utilizes the highest WIMAX 802.16d transmission range (31 miles). All BS(s) in the BSL are linked to the ACU via fast speed connection (optical fiber) and cooperate in transmission/reception and beamforming by virtue of the MIMO processing at the ACU.

18. The architecture of claim 9 wherein the ACU is comprised of a MIMO-OFDM processing unit which does the MIMO modulation/demodulation algorithms (SVD and water pouring algorithm or Schmidt-Gram in combination with power method), MIMO encoding/decoding, Channel space time training, channel estimation, beamforming algorithm and TDD calibration and synchronization etc. ACU acts as interface (gateway) to external networks. It also has a high speed optical fiber feedback link with the JACU at the APL for the purpose of TDD, CSI and perfect calibration. The ACU can achieve good collision avoidance protocol by controlling the antenna/beam selection switch.

19. The architecture of claim 9 wherein an AP comprises of a MIMO-OFDM transceiver unit comprised of a number of antenna elements N and N-OFDM streams. It operates as average capacity unit utilizing the 802.16d 8-10 miles transmission range. The AP basic functions include but are not limited to: RF interface, joint beamforming with AP(s) in the global (integrated) systems (BS-AP and AP-EU), and beam selection, virtual array partitioning and beamforming in the local (AP MEA) array. More than one AP at a time communicates with a BS or more, the number is such that the channel matrix in not rank deficient. The AP(s) are linked optical fiber to the JACU which performs

of switching communication streams between BSL and EUL.

20. The architecture of claim 9 wherein the numerical relation between the M elements in a BS and N elements in an AP; using beam switching at BS and AP-in addition to antenna switching at ACU and JACU- the rule is such that for M elements in a BS there are corresponding M beams from different AP(s) and for N elements in an AP there are corresponding N beams from different BS(s).

21. The architecture of claim 9 wherein the method of partitioning an antenna array [PCT/US/2006/022315] is implemented at the BS units and the AP units to create local artificial diversity between AP-BS transmit receive pairs and to exploit diversity gain as well.

22. The BS of claim 17 or the AP of claim 19 in combination with the method in claim 21 wherein local beam switching with butler matrix of the locally created virtual arrays is used.

23. The architecture of claim 9 wherein the AP(s) performs dual communication simultaneously with BS(s) in MIMO-SU and with EU(s) in MIMO-MU with TDMA.

24. The architecture of claim 9 wherein the AP(s) performs dual communication simultaneously with BS(s) in MIMO-SU and with EU(s) using sectored architecture.

25. The architecture of claim 9 wherein the positions of the AP(s)/BS(s) transmit/receive pairs are fixed; the feedback line is replaced by look up tables of channel state numbers and auto calibration weights for specific angles of arrival/departures and spatial signatures. These tables are developed through campaigns and afterwards stored in the processing units.

26. The method of claim 1 wherein general applications to terrestrial satellite networks, VSAT and BGAN are indicated in paragraphs [030], [031], [032].

U