IL111340A - Frequency hopping multiple access communication system - Google Patents

Description

111340/2 FREQUENCY HOPPING MULTIPLE ACCESS COMMUNICATION SYSTEM POWERSPECTRUM TECHNOLOGY LTD C: 20385 20385NEW.SPC DZ-23085X 03JAN96 A FREQUENCY HOPPING MULTIPLE ACCESS COMMUNICATION SYSTEM FIELD OF THE INVENTION The present invention relates to apparatus and method for providing voice and data communications. More specifically, it relates to apparatus and method for providing a frequency hopping multiple access communication system.

BACKGROUND OF THE INVENTION Multiple access communications systems are capable of providing multiple communications at the same time using the same system resources. These systems utilize various communications protocol and various system architectures. One protocol is time division multiple access wherein users communicate on a shared channel at different times. Another protocol is frequency division multiple access wherein separate frequency channels are allocated to mobile radio terminals. The system architecture commonly utilized is a cellular type architecture wherein there are many base stations that provide communication channels to many radio terminals.

The existing multiple access communication systems all have various drawbacks. By way of example only, many rely on a hardware intensive architecture that is costly to implement and also costly to operate. These high costs result in a higher cost to customers. The existing multiple access communication systems also have severe limitations on system capacity or, stated differently, require a great deal of spectrum to efficiently operate. The existing multiple access communication systems also have limitations on the quality and the types of the communications services provided, particularly when the system is fully loaded.

Thus, a communication system that offers low cost implementation and operation along with improved communication characteristics is needed.

SUMMARY AND OBJECTS OF THE INVENTION The present invention generally relates to method and apparatus for providing multiple access communications. In one aspect of the present invention, method and apparatus for handing off communications between a communicating party and a mobile communication radio which is moving from a first sector to an adjacent sector in a communication base station having multiple sectors is provided. In accordance with this aspect of the present invention, a mobile radio detects synchronization information from the first sector and searches for synchronization information from sectors adjacent to the first sector. The mobile radio determines when the synchronization information from the adjacent sector has stronger reception then the synchronization information from the first sector and then requests a hand off from the base station. The base station enables a three way communication link between the mobile radio in the first sector, the radio in the adjacent sector and the communicating party. Then, voice activity from the base station to the mobile radio is monitored and downlink communications from the base station to the mobile radio are handed off from the first sector to the adjacent sector when voice inactivity is detected. Also, voice activity from the mobile radio to the base station is monitored and uplink communications from the mobile radio to the base station are handed off from the first sector to the adjacent sector when voice inactivity is detected.

It is an object of the present invention to provide a radio communication system.

It is another object of the present invention to provide a system and process of providing a frequency hopping radio communication system.

It is a further object of the present invention to provide a system and method of providing a time hopping radio communication system.

Another object is to provide apparatus and method for handling subscriber units attempting to communicate in fringe areas .

Yet another object is to provide apparatus and method for providing service to subscriber units located at the foot of the base station.

Yet a further object is to provide apparatus and method of measuring the quality of service provided by a frequency hopping communication system.

It is also an object to provide apparatus and method that provide talk around capabilities.

It is another object to provide apparatus and method that acquire and track the communication channels of the present invention.

Another object of the present invention is to provide method and apparatus for handing off communications between sectors .

It is a further object to provide apparatus and method of automatic frequency control.

It is also an object to provide apparatus and method that provide linear power amplification of the signals to be transmitted.

Another object is to provide apparatus and method for automatic gain control.

These and other objects are further described in the description of the preferred embodiment that follows.

BRIEF DESCRIPTION OF THE DRAWINGS The present invention will be understood and appreciated more fully from the following detailed description, taken in conjunction with the drawings in which: Fig. 1 represents the communication system of the present invention; Fig. 2 illustrates the preferred common air interface links between a base station and the subscriber units; Fig. 3A illustrates the preferred rules of transmission for traffic, - control and access channels in each of the three sectors of the preferred embodiment of the present communication system; Fig. 3B shows one way of time and frequency hopping multiple inputs and Fig. 3C illustrates a sequence generator for generating the hopping sequences of the present invention; Fig. 4 shows a preferred time slot format; Fig. 5 illustrates the preferred timing offset between uplink and downlink transmissions; Fig. 6 illustrates a preferred sync-label slot format; Fig. 7 illustrates the transmission of the SLS; Fig. 8 is a block diagram of a base station; Fig. 9 is a block diagram of a sector unit in a base station ; Fig. 10 illustrates a block diagram of the interconnection of the frame processor and the transmit (Tx) processing unit; Fig. 11 illustrates the HDLC interconnections between the frame processing unit, the Tx processing unit, the Rx processing unit and the sector computer; Fig. 12A illustrates a frame processing unit and Fig. 12B illustrates the block diagram of a quad frame board in the frame processing unit; Figs. 13A and 13B illustrate a Rx processing unit and a Tx processing unit, respectively.

Fig. 14 is a block diagram of a micro sector unit in a base station; Fig. 15 illustrates a preferred embodiment of a subscriber unit; Fig. 16 illustrates a preferred embodiment of the RF portion of the subscriber unit; Fig. 17 illustrates a preferred method of synthesizing frequencies ; Fig. 18 illustrates a preferred embodiment of a modem of the subscriber unit; Fig. 19 illustrates a preferred embodiment of an ASIC in the modem of the subscriber unit; Fig. 20 illustrates a preferred controller in the subscriber unit; Fig. 21 illustrates a preferred voice package processor (VPP) in the subscriber unit; Figs. 22A and 22B illustrate a preferred service board found in the subscriber unit; Figs. 23 and 24 show, the preferred signal processing performed when transmitting and receiving signals on traffic channels, respectively; Figs. 25 to 31 show various error coding schemes; Figs. 32 to 38 illustrate various steps used in other processes of the present invention; Fig. 39 is a simplified block diagram illustrating the overall structure of the software components of the system of Fig . 1 ; Fig. 40 is a simplified block diagram illustrating the structure of element 700 of Fig. 39 in greater detail; Fig. 41 is a simplified block diagram illustrating the structure of element 706 of Fig. 40 in greater detail; Figs. 42A and 42B are simplified block diagrams illustrating the structure of element 708 of Fig. 40 in greater detail ; Figs. 43A and 43B are simplified block diagrams illustrating the structure of element 714 of Fig. 42 in greater detail ; Fig. 44 is a simplified block diagram illustrating the structure of element 724 of Fig. 42 in greater detail; Fig. 45 is a simplified block diagram illustrating the structure of element 718 of Fig. 42 in greater detail; Fig. 46 is a simplified block diagram illustrating the structure of element 720 of Fig. 42 in greater detail; Fig. 47 is a simplified block diagram illustrating the structure of element 716 of Fig. 42 in greater detail; Figs. 48A and 48B are simplified block diagrams illustrating the structure of element 702 of Fig. 39 in greater detail; Figs. 49 - 51 are simplified block diagrams illustrating the structure of a microsite; and Fig. 52 is a simplified block diagram illustrating the structure of a remote sector.

Attached herewith are the following appendices which aid in the understanding and appreciation of one preferred embodiment of the invention shown and described herein: PS SPECIFICATION LIST PS 1. 0 SYSTEM PS 1. 1 System Specs PS 1. 4 Common Air Interface PS 1. 5 VPP PS 1. 7 ICD PS 2. 0 SYSTEM : SOFTWARE SPECS PS 2. 1. 00 Power Spectrum System PS 2. 1. 01 Administration PS 2. 1. 02 Connection MGR 1 PS 2. 1. 03 Connection MGR 2 PS 2. 1. 04 Debug PS 2. 1. 05 Data Base MGR PS 2. 1. 06 Dispatch MGR PS 2.1.07 Network MGR PS 2. 1. 08 Operator Application PS 2. 1. 09 PABX I/F 1 PS 2. 1. 10 PABX I/F 2 PS 2. 1. 11 SU MGR PS 2. 1. 12 Telephone MGR PS 2. 1. 13 Voice Mail PS 2. 1. 15 BIT MGR PS 2. 1. 16 Base S/W PS 2. 1. 17 Inter Communication Unit PS 2. 1 18 Inter Communication Frame PS 2. 1. 19 Initiator/Terminator PS 2. 1. 20 Initiator/Terminator MGR PS 2. 1. 21 UTL_SRS . book . B8665 PS 2. 1. 30 Physical Layer PS 2. 1. 31 Slot Processor PS 2. 1. 32 Frame' Processor PS 2. 1. 34 Subscriber Control PS 2. 1. 35 Subscriber DSP PS 2. 1. 38 Bridge Interface Board - HCM PS 2. 4 Interface Design Document ADDITIONAL DOCUMENTATION RS Remote Sector Functional Requirements for Geonet Services Data System Requirements for Geonet System PowerSpectrum Processing and Servicing PowerSpectrum Specification Air Interface Me Provisions for Data Data Provisions for the TSU Hand_Off/DTO Heart_Beat/DTO Processes G D Mechanism Time Alignment and Ranging Update of Process Start Dispatch Registration Processes Registration Processes REG-1 MARQ to TSUC Network and Management Control Algorithms Document DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS Fig. 1 illustrates the system 1 of a preferred embodiment of the present invention. The system 1 includes a base station 10', a plurality of subscriber units 12, three sectors 14, 15 and 16, a microsite 18 and a remote site 20.

In accordance with the present invention, the base station 10 establishes a communication link between a user on one subscriber unit 12 and one or more other users on other subscriber units 12. The base station 10 can also establish connections between one or more subscriber unit 12 and the Public Switch Telephone Network (PSTN) .

In accordance with a preferred embodiment of the present invention, the communication system 1 is divided into sectors. While Fig. 1 illustrates three sectors 14 to 16, the preferred number of sectors utilized will depend mainly on the geographic location of the communication system 1 and on the number of subscriber units 12 which the system 1 needs to support.

Common Air Interface The communication links between the base station 10 and subscriber units 12 is referred to as the common air interface. Fig. 2 illustrates a set of preferred communication channels in the common air interface. As shown the communication channels include a plurality of traffic channels (TCHs) , one or more control channels (CCHs) and one or more access channels (ACHs) . All of these channels are present in each sector 14 to 16. The TCHs operate in the uplink (transmissions from subscriber units 12 to the base station 10) and in the downlink (transmissions from the base station 10 to the subscriber units 12) . The CCHs and the ACHs, however, operate only in one direction — the CCHs in the downlink and the ACHs in the uplink.

In the communications system 1 of the present invention, these channels are transmitted over a plurality of carrier frequencies. The number of carrier frequencies (or channels) utilized by the system 1 depends mainly on the available frequency spectrum and on the loading of the system 1. Each carrier frequency is preferably reused in each sector 14 to 16.

In one embodiment, in each sector 14 to 16, the system 1 uses ten carrier frequencies to define ten uplink channels and uses a different ten carrier frequencies to define ten downlink channels. In this embodiment, each uplink and downlink channel — whether TCH, CCH or ACH — operates within a channel bandwidth allocation of 25kHz. Also, in this embodiment, the uplink band of frequencies is contiguous as is the downlink band of frequencies, although such operation is not necessary to the present invention. In each sector 14 to 16, nine out of the ten available uplink carrier frequencies are utilized to implement nine uplink TCHs and nine of the ten available downlink carrier frequencies are utilized to implement nine of the ten downlink TCH transmissions. In each sector 14 to 16, the remaining uplink carrier frequency is utilized to transmit a single ACH while the remaining downlink carrier frequency is utilized to transmit a single CCH.

Each of the channels illustrated in Fig. 2 — the TCHs, the CCHs and the ACHs — carries predefined information. The TCHs transmit voice information, data information and inband overhead control signals between the base station 10 and the subscriber units 12. The CCHs transmit timing and control signals from the base station 10 to the subscriber units 12. At least one CCH is preferably transmitted perpetually by each sector unit 14 to 16. The ACHs transmit status and operational requests from the subscriber units 12 to the associated sector unit 14 to 16 in the base station 10.

Referring initially to Fig. 3A, several preferred channel transmission rules in the sectors 14 to 16 will be discussed. The first rule concerns the transmission of information over the TCHs. As previously described, in accordance with one preferred embodiment of the present invention, the TCHs are defined over nine uplink and nine downlink carrier frequencies. The transmission of information over a TCH is preferably via frequency hopping so that at a first time, a first block of information is transmitted on a first carrier frequency while at a second time, a second block of information is transmitted on a second carrier frequency, and so on. The transmitted information, therefore, is preferably hopped from carrier frequency to carrier frequency.

It is also preferred that each TCH be constructed of a periodic sequence of fixed, continuous and non-overlapping time slots. In accordance with a preferred embodiment, there are three time slots, but a larger or smaller number can easily be implemented. Some of the slots will be active, meaning that there is a transmission during that slot. Other slots will be passive, meaning that there is no transmission during that slot.

Referring to Fig. 3B, an example of frequency and time hopping of inputs from multiple users by the base station 10 is illustrated. In Fig. 3B there are five users, namely users A to E. User A has four blocks of information — Al, A2 , A3 and A4 — which are queued for transmission. Similarly, user B has four blocks of information — Bl, B2 , B3 and B4 — that are queued for transmission, user C has four blocks of information — CI, C2 , C3 and C4 — that are queued for transmission, user D has four blocks of information — Dl, D2 , D3 and D4 -- that are queued for transmission, and user E has four blocks of information — El, E2 , E3 and E4 — that are queued for transmission.

Referring to the right side of Fig. 3B, the available channels — in this case ten -- which are defined by the carrier frequencies f -^ to f]_0' are illustrated. Each of the ten illustrated channels are constructed of continuous time slots. There are four periods of transmission (designated I, II, III and IV) illustrated and each transmission period is divided into three time slots A, B and C.

The queued information on the left side of Fig. 3B is processed by assigning each block of information from each user to a time slot and then to a carrier frequency. For example, user A's information blocks, Al, A2 , A3 and A4 are assigned to the time slot IB and to the channel f3 , to the time slot lie and to the channel fg, to the time slot IIIA and to the channel f9 , and to the time slot IVB and to the channel f2, respectively. The blocks of information to be transmitted, therefore, are time hopped between the available time slots. They are also frequency hopped between the available channels. The assignment to time slots and to channels is preferably made ■ in accordance with certain rules which will be discussed later. The queued information blocks for users B, C, D and E are similarly time and frequency hopped, as illustrated.

The frequency hopping is preferably done in accordance with a predefined sequence, which can be modified as necessary. It is preferred, to select the sets of hopping sequences in an individual sector, for example, sector 14, such that during a selected time slot each of the carrier frequencies only being used by a single channel. Stated another way, no two channels within a sector employ the same carrier frequency at the same time. Selection of hopping sequences in accordance with this rule eliminates interference within the sector 14. This pr&ferred transmission rule and the generation of the sequences is more fully described in co-pending United States : patent number No. 5,408,496 granted April 18, 1995 (and in co-pending Israel '"'application number 103,620, filed 3 November 1992) which is hereby incorporated herein by reference.

The second preferred transmission rule also relates to the transmission of TCHs . In accordance with the second rule, the set of frequency hopping sequences in one sector are selected such that no channel in that set of sequences employs the same carrier frequency at the same time as more than a predetermined number of channels in another set of frequency hopping sequences in an adjacent sector. The predetermined number of channels is the minimum number of channels possible and in a preferred embodiment is one. This transmission is discussed in greater detail in co-pending United States patent number 5,408,496 .which has ■been incorporated herein by reference.

Referring to Fig. 3C, a preferred sequence generator is illustrated. A sequence is chosen for the first sector. The sequences for the second sector are generated by a left cyclic shift of each line of the sequences of the first sector by one location. The sequences for the third sector are also generated by a left cyclic shift of each line of the sequences of the second sector by one location.

Fig. 3A also describes the preferred transmission rules for CCHs which transmit timing and control signals from the base station 10 to the subscriber units 12. As already described, there is a single CCH in each sector 14 to 16. In accordance with a preferred embodiment, the single CCH is perpetually transmitted over one of the plurality of carrier frequencies. Therefore, the CCH is preferably not frequency hopped. Further, the CCH in each sector 14 to 16 is preferably assigned one of the three time slots and is transmitted only in that time slot. Therefore, in a preferred embodiment, the CCH in sector 14 is transmitted during time slot A, the CCH in sector 15 is transmitted during time slot B and the CCH in sector 16 is transmitted during time slot C.

Fig. 3A also describes the preferred transmission rules for ACHs. In each sector 14 to 16, most of the ACHs are transmitted in a slotted ALOHA format. It is preferred to utilize a variant of the stabilized slotted. Typically, a portion of the ACHs is in a format similar to the CCH.

As already stated, each channel is constructed of continuous time slots. Fig. 4 illustrates a preferred format of an active time slot 20. The slot 20 covers 2.22 msec. and includes a total of 41 symbols. Each symbol, therefore, is 54 micro seconds long. Each symbol may consist of two bits of information. Two of the symbols, the first and last ones, are left inactive as guards to protect against time shifted time slots. When the time slot is active, therefore, the middle 39 symbols are active and used to transmit information.

It is preferred to transmit the uplink slots 21 from the subscriber unit 12 to the base station 10 offset in time from the transmission of downlink slots 22 from the base station 10 to the subscriber units 12, as shown in Fig. 5. The time offset is preferably 1.11 msec. Offsetting the uplink transmissions from the downlink transmissions allows the system of the present invention to be compatible with either full duplex (receive and transmit simultaneously) or half duplex operation.

As stated previously, timing and control information for the system 1 is transmitted over a CCH. One of the CCH messages contains synchronization information and is transmitted in sync/label slots (SLS) . Fig. 6 shows a preferred format of a SLS 23. The SLS 23 includes a sync 24 consisting of 20 symbols, a label 25 consisting of 19 symbols and guard symbols 26 and 27 at the front and the end of the SLS 23. The sync 24 is a deterministic synchronization code. The subscriber units 12 detect the sync 24 when it is transmitted and synchronize their operation accordingly. The label 25 is preferably constructed^- of eight symbols that indicate which communication site transmitted the SLS, four symbols indicating which sector within the site the subscriber unit is located in, six symbols indicating the index of the received SLS 23 within a superframe of information and 1 symbol indicating whether coherent or differential modulation is being utilized.

The transmission of the SLS 23 is illustrated in Fig. 7. As illustrated, each sector 14 to 16 is assigned its own time slot. The SLSs are transmitted on the average, every 72 time slots in a staggered manner so as to ensure the reception of at least one SLS in one's own operating time slot. The entire pattern has a period of 216 time slots. Information about the base station 10 is also transmitted on the CCH.

It is preferred to have two modes of demodulation operation. The first mode of operation is coherent demodulation and the second mode of operation is differential modulation. Coherent demodulation offers improved performance when the received signal has a stable carrier phase, but requires good synchronization to the phase of the received, yet unmodulated, signal. It is anticipated that some channel conditions — such as severe fading — may be such that coherent demodulation may be erratic. It is, therefore, preferred to utilize differential demodulation as well. The mode of operation will be initially selected for individual communication sites. The base station 10 instructs the subscriber units 12 which mode of operation is being used by setting a bit in the SLS 23.

The Base Station The main task performed by the base station 10 is to connect subscriber unit users 12 with a PSTN or with another subscriber unit 12. The base station 10 provides various voice services to the users, including basic telephony, voice mail, group dispatch calls and individual dispatch. The basic telephony services include incoming phone calls from a PSTN, outgoing phone calls outside the base station 10 area via a PSTN and phone calls between subscribers. The preferred features provided for telephony services includes speed dialing based on a phone number library stored in the subscriber unit 12, fleet phone numbers library, call waiting, call forwarding, emergency call, display of caller phone number, camp-on capability, prescribed telephony privileges per subscribe unit 12 and answering modes that include auto answer and ringing. The features of the group dispatch services includes emergency dispatch and camp-on capability on system resources and on group dispatch. Individual dispatch can be provided via full duplex and via half duplex.

The base station 10 also provides data services, including group messaging, individual messaging and virtual circuit. For individual messaging, basic individual messaging services, special delivery and registered delivery are provided. Also provided are: warnings to the sender; guaranteed delivery; unique identification of messages, a dropped message log file; a message receipt time stamp; two levels of message priority; delayed delivery; an address distribution list and an alternate destination address. The virtual circuit provides a connection oriented, packet switched type of service between two users. It is preferred to measure inactivity time for each virtual circuit. The base station 10 also provides fax services. Another basic task performed by the base station 10 is to manage a date base of the subscriber unit 12 users.

A preferred embodiment of the base station 10 is illustrated in Fig. 8. The base station 10 includes a first sector unit 30, a second sector unit 31, a third sector unit 32, a microsector unit 34, a redundant sector 35, a PABX 36, a voice mail unit 38, a central frequency source unit 40, an administration computer 42, a central controller 44, a data base server 46, a local administrator computer 48, a terminal server 50, an ethernet local area network 52, a power supply 54 and data computers 55.

The sector units 30 to 32 establish the previously discussed communication channels within the sectors 14 to 16, respectively. The micro sector unit 34 establishes communication channels with the microsite 18. The redundant sector 35 provides redundant communication channels for the various sector units. Communication with the PSTN is provided via the PABX 36.

Referring to Fig. 9, a block diagram of a sector unit 30 to 32 is illustrated. The sector unit 30 includes antennae 56 and 57, antenna boxes 58 and 59, wideband amplifier units 60 and 61, each of which include an input combiner 62, a preamplifier 63, a power divider 64, amplifiers 65 and an output combiner 66, multicoupler units 67A and 67B, a receive (Rx) processing unit 68, a transmit (Tx) processing unit 70, a frame processing unit 72, a sector controller 74 and a power supply 76.

Both antennas 56 and 57 are utilized during a receive cycle and during a transmit cycle. During a transmit cycle, the antenna 57 transmits signals provided by the Tx processing unit 70 through the wideband amplifier 69 and the antenna box 58 into a sector while the antenna 56 transmits signals provided by the Tx processing unit 70 through the wideband amplifier 61 and the antenna box 59. During a receive cycle, the antennae 56 and 57 both receive signals that are provided to the multicouplers 67A and 67B, respectively, before being sent to the Rx processing unit 68. Thus, two complete receive paths are established to provide spatially independent received signals. The antennae 56 and 57 are preferably placed more than several wavelengths from each other. The best received signal is selected from the two and then routed for frame processing.

Referring to Fig. 10, a block diagram of the interconnection between the frame processor unit 72 and the Tx processing unit 70 is illustrated. The frame processing unit preferably includes forty (40) FPUs, while the Tx processing unit 70 preferably includes ten (10) TPUs. Each FPU is connected by the sector controller to handle one conversation at a time. Thus, as illustrated, up to forty conversations can be handled by the forty FPUs. When an FPU is instructed to handle a conversation, the sector controller provides the FPU with a key that specifies the sequence of frequencies to be utilized. The key can also specify the time slot to be utilized. Each FPU is connected to each TPU in the Tx processing unit 70 through a modified HDLC bus. Each TPU provides signals to be transmitted on a predefined frequency (channel) . The FPU sends the part of the conversation to be transmitted to the appropriate TPU on the HDLC bus in accordance with the key received from the sector controller. The time slot of transmission is determined in accordance with the time position of the signal to be transmitted on the HDLC bus. The FPU properly places the signal to be transmitted in the proper time sequence on the HDLC bus in accordance with the key provided by the sector controller.

Fig. 10 also represents the interconnection between the frame processor unit 72 and the Rx processing unit 68. The Rx processing unit has ten RPUs (represented by the TPUs in Fig. 10) , each of which receives signals from the appropriate sector on a predefined frequency (channel) . The sector controller instructs the RPU which FPU should receive the signal through a key. The RPU then sends the information received in the proper timing on the HDLC bus to the appropriate RPU where the signal transmitted is reconstructed. Fig. 11 illustrates the HDLC interconnections in greater detail.

Referring to Fig. 12A, a preferred frame processing unit 72 is illustrated. The frame processing unit 72 includes a bridge interface board 80, a dual Tl board 82 and ten quad frame boards 84. The dual Tl board 82 provides a communication interface to the PABX 36. The bridge interface board 80 provides an interface to the HDLC bus. Each one of the quad frame boards 84 includes four FPUs so that a total of forty FPUs are provided as previously discussed.

The quad frame board 84 is illustrated in greater detail in Fig. 12B. Each board 84 includes a host CPU 86. Each board also includes four FPUs, each of which includes a digital signal processor 88, a vocoder 90, buffers 92 and 94, a decoding PAL 96, an oscillator 98, a test buffer 100, a boot ram 102, a viterbi decoder 104, data ram 106 and program ram 108. The host CPU 86 controls the communications on the HDLC bus as well as the overall functioning of each FPU. Each FPU processes the information to be transmitted or received. On the transmit side, the processing includes compressing the information with a vocoder, error correction coding and interleaving the signal. On the receive side, the processing includes the reverse steps.

Referring to Figs. 13A and 13B, the Rx processing unit 68 and Tx processing unit 70, respectively, are illustrated. The Rx processing unit 68 preferably includes bridge interface boards 110 and 112, Rx RF/IF boards 114 and 116, digital Rx boards 118. and 120 and reference generator 121. The bridge interface boards 110 and 112 provide an interface to the Tx processing unit 70, where an interface to the frame processing unit 72 is provided. The signals being received are provided to the Rx RF/IF boards 114 and 116 by the multicouplers unit 67A and 67B, respectively, where the signals are converted to digital form. Once in digital form, the signals are provided to the digital Rx boards 118 and 120, where the signals are put in proper form for transmission to the frame processor 72.

Referring to Fig. 13B, the Tx processing unit 70 is similar to the Rx processing unit 68, but is configured to send signals in the opposite direction. The Tx processing unit 70 includes bridge interface boards 122 and 124, Tx RF/IF boards 126 and 128, digital Tx boards 130 and 132, and reference generator 133. The bridge interface boards 122 and 124 provide an interface for the Tx RF/IF boards 126 and 128 to the frame processing unit 68. The bridge interface boards 122 and 124 also provide an interface to the Rx processing unit 68, as shown. The digital Tx boards 126 and 128 collect the signals to be transmitted on the various channels and properly format these signals. The Tx RF/IF boards 126 and 128 convert the signals to be transmitted to analog form for transmission.

Referring back to Fig. 9, the sector controller 74 provides control of the various functions performed by a sector and is controlled by the central controller 44. Control and data signals are exchanged between the sector controller 74 and the central controller 44 via the LAN 52. In addition to controlling the processing that is performed by the sector 30 to 32, the sector controller 74 also keeps a log book of all the active users under supervision. The sector controller 74 provides this information to the central controller 44 via the LAN 52.

Referring back to Fig. 8, the description of the base station 10 will be completed. Each of the sector units 30 to 32 and the micro sector unit 34 is interfaced to the PABX 36. The PABX 36 provides the connection of the system 1 to the public switch telephone network (PSTN) . The interface is via a standard 2Txl connection. The PABX 36 also provides three way conferencing, routing, least cost routing of long distance calls, voice mail interfacing, dispatch bridging, user services support and metering functions.

The voice mail unit 38 provides voice mail capability. The central frequency source unit 40 provides sector and timing references to the sector units 30 to 32 and to the micro sector unit 34. The frequency source is preferably generated by a rubidium atomic reference.

The administrative computer 42 tracks the configuration grouping, tracks administration activities, performs network management, performs bit management and performs the system initialization. It is preferably implemented with a Sun Sparc-station.

The central controller 44 provides various functions including call management, dispatch management, control of the PABX 36, voice mail interfacing, operational mode management, subscribers management, call management, billing information and reports generation. It is preferably implemented with a 486 PC compatible computer. The data base server 46 stores user data concerning user rights, status, calls and airtime. It also provides basic data base management and services to all data base clients, such as the local operator, fleet administrators and remote operators. The local administrator computer 48 provides maintenance and operational control of the base station 10. An ethernet local area network (LAN) 52 is provided to enable communications between the various components connected to the network.

The micro sector unit 34 is illustrated in greater detail in Fig. 14. The micro sector unit 34 provides communication channels to the microsites 18. The micro sector 34 includes one or more microwave transceivers 134, a video processing unit 136, a micro Tx processing unit 138, a micro Rx processing unit 140, a slot selector unit 142, a frame processing unit 144, a sector controller 146 and a power supply 148. ' The receive and transmit signal processing is performed by the micro Tx processing unit 138, by the micro Rx processing unit 140, by the slot selector unit 142, by the frame processing unit 144 and by the sector controller 146. The processing steps performed by the micro sector 34 are the same as those performed by the sectors 30 to 32 as previously discussed.

Subscriber Unit Fig. 15 is a block diagram of the subscriber unit (SU) 12 of the present invention. Functionally, the SU 12 preferably supports the same features supported by the base station 10. As previously discussed, these features include, but are not limited to, telephony interconnect, two-way radio, and data communications. The telephony interconnect features include incoming calls, outgoing calls, call waiting, call forwarding, call disconnect, call flip-flop, switching to dispatch and back while on call, dialing while conversing, last number redial and speed dialing. The two-way features include incoming dispatch and dispatch initiation.

As shown in Fig. 15, the SU 12 includes a first antenna 202, a second antenna 204, a radio unit (RU) 206, a baseband unit (BBU) 208, a service board (SB) 210, a GPS interface 211 and a man machine interface (MMI) 212. The RU 206 includes a duplexer 213, a receiver channel 214, a diversity receiver channel 216, a gain and frequency control unit 218, a transmitter 220, a synthesizer 222 and a gain control unit 224. The BBU 208 includes a modem 226, a controller 228, a voice processing package (VPP) 230 and a MMI interface 232.

The two antennae 202 and 204 establish the previously discussed communication channels with the base station 10. They are preferably 14 inch, stainless steel, high performance collinear antennae which are magnetically secured to a vehicle. They preferably operate in the frequency range of 890 to 950 MHz, although tuning to any desired frequency is possible. The two antennae 202 and 204 are preferably omnidirectional and have linear vertical polarization, a free space gain of 3 dBi and a maximum VSWR of 1.5:1.

The SU 12, when transmitting to the base station 10, transmits only on the first antenna 202. When the SU 12 receives transmissions from the base station 10, however, both the first and second antenna 202 and 204 are utilized to achieve space diversity. As with the base station 10, the best signal is selected for processing.

Referring to Fig. 16, a more detailed block diagram of the transmitter 220, the receivers 214 and 216 and the synthesizer 222A and 222B is illustrated. The transmitter 220 received I and Q inputs from the modem 226. The I and Q inputs are amplified by amplifiers 234 and 236, respectively, and then filtered by low pass filters 238 and 240, respectively. The filtered I and Q signals are then modulated by a modulator 242. The modulator 242 is supplied with a hopping oscillator signal, TxLO, from the synthesizer 222A so that the signal transmitted by the transmitter 220 is frequency hopped.

The modulated signal is then controllably attenuated by an Up attenuator 244, filtered by a high pass filter 246, amplified by amplifiers 248 and 250, down attenuated by attenuator 252 and filtered by low pass filter 254 before being supplied to the duplexer 213 for transmission by the antenna 202. As can be seen, the gain of the transmitted signal is controlled by the attenuators 244 and 252. The attenuator 244 is controlled by the signal LEVEL CONTROL which is received from the gain control unit 224. The gain control unit 224 receives its inputs (EN5, DATA and CLK) from the modem 226.

The antenna 202, in addition to transmitting signals from the transmitter 220 to the base station 10, also receives signals which are transmitted by the base station 10. Those received signals are transmitted through the duplexer 213 to the receive channel 214. The received signal is amplified with a low noise amplifier 256, filtered by a bandpass filter 258 and then downconverted by a mixer 260 to IF.

The received signal can be a frequency hopped signal. Where appropriate, therefore, the mixer 260 downconverts the received signal by mixing the received signal with an oscillating signal from the synthesizer 222B which is also hopping.

The downconverted signal is then filtered by a bandpass filter 262 and then amplified by a variable amplifier 264. The variable amplifier 264 is gain controlled in accordance with a signal, AGC, received from the gain & frequency control unit 218.

The signal transmitted from the base station 10 to the SU 12 is also received by the second antenna 204. The received signal is then processed by the receiver 216. First, the received signal is filtered by a band pass filter 272 and then it is processed by a low noise amplifier 274, by a filter 276, by a mixer 278, by a mixer 280, by a gain controlled amplifier 282, by a mixer 284, by a filter 286 and by an amplifier 288 in a similar fashion to the processing performed in the receiver 214. Note, however, that gain control of the diversity channel is accomplished by a separate control signal, AGC(D).

The synthesizer 222A and 222B generates the signals necessary to modulate the signal transmitted by the transmitter 220 and to downconvert the signal received by the receiver channels 214 and 216. A fixed frequency generator 296 generates a signal having a frequency of 14.4 MHz. This signal is provided to a transmit hopping synthesizer 294. The transmit hopping synthesizer 294 generates an output signal with a frequency that varies from 757.8 MHz to 797 MHz in accordance with control signals, TxHOP CONTROL, which are provided by the modem 226. The output of this synthesizer is output to a divider 294 where the frequency of the synthesized signal is divided by eight. Then the signal is filtered by high pass filter 296 before being supplied to one of the inputs of a mixer 298.

The other input to the mixer 298 is supplied by a fixed synthesizer 300 which synthesizes a frequency of 801.29375 MHz in accordance with an input from an oscillator (TCXO) 302. The TCXO 302 is controlled by a control signal, AFC, which is generated by the modem 226. The control signal AFC is varied to keep the frequency of transmission constant. The output of the mixer 298 therefore, is a hopped frequency which is filtered by a bandpass filter 304 before being input to the modulator 242 where it is used to modulate the I and Q signals that have been provided by the modem 226.

An output from the reference 290 is also supplied to a receive (Rx) hopping synthesizer 306. This synthesizer 306 generates a signal that is hopped in the frequency range of 882.3 MHz to 902.2 MHz. The frequency hopping is controlled by control signal Rx HOP CONTROL generated by the modem 226. The output of the Rx hopping synthesizer 306 is supplied to a divider 308 and then to a high pass filter 310 before being supplied to an input of a mixer 312. The other input of the mixer 312 is supplied by the fixed synthesizer 300. The output of the mixer 312 is filtered by a band pass filter 314 and results in a signal that ranges in frequency from 1021.86875 MHz to 1026.84375 MHz. The signal is then split into two signals by a splitter 316. One of the split signals is supplied to the downconverter 260 in the first receiver channel 214. The other signal from the splitter 316 is supplied to the downconverter 278 in the diversity channel 216.

The reference 290 also supplies a reference frequency to an IF synthesizer 318. The IF synthesizer 318 generates a signal having a frequency of 86.38875 MHz. This signal is split by a splitter 320 and then supplied to downconverters 266 and 284 in the receiver channels 214 and 216, respectively.

Referring to Fig. 17, the preferred method for generating the hopping oscillators is illustrated. A frequency reference 310 drives a numerical controlled oscillator 311. The two sine and cosine outputs (in quadrature) from the numerical controlled oscillator drive a single sideband (SSB) mixer 312. An additional RF input is provided from a voltage controlled oscillator 313 through a directional coupler 314 at frequency fouf The outPut of tne mixer 312 at frequency fout ~ ½CO is down divided by N using a frequency divider 315. A phase lock loop, formed by VCO 313, SSB mixer 312, divider 315, phase detector 316, and loop filter 317, is used to lock the VCO onto an output frequency such that fR * N = fOU†- - ½αο· Coarse frequency is set by fR and N and fNC0 is used for fine and fast frequency hopping.

Fig. 18 illustrates the circuitry of the modem 226. The modem includes a digital to analog converter (DAC) 322, an analog to digital converter (ADC) 324, a converter interface (CNVR) 326, an ASIC 328 and a digital signal processor 330. The DAC 322, during the transmit function, receives I and Q signals from the ASIC 328 and converts those signals to analog form before supplying them to the transmitter 222. The ADC 324, during the receive function, receives a signal IF from the receiver channel 214 and a diversity signal IF-d from the diversity receiver channel 216. It converts these signals to digital form and then supplies them to the ASIC 328 for further processing. The CNVR 326 provides the control signals for automatic gain control of the first receiver channel 214 (AGC) , for automatic gain control of the diversity receiver channel 216 (AGC-d) and for automatic frequency control of the TCXO generator 302. The CNVR 326 also receives the input from the temperature sensor in the radio unit 206. These signals are either received from or transmitted to the ASIC 328.

The ASIC 328 is illustrated in greater detail in Fig. 19. The ASIC 328 includes a transmit (Tx) port interface 332, a receive (Rx) port interface 334, a CNVR interface 336, a Tx control circuit 338, a PLL control 340, a lock indicator 342, a DSP bus interface 344, a Viterbi decoder 346, a timing & interrupt controller 348 and a controller interface 350. The Tx port interface 332 provides digital I and Q signals to the DAC 322. The Rx port interface 334 receives the digital signals from the receiver channels 214 and 216 through the ADC 324. The CNVR interface 336 provides the control signals AGC, AGC-d and AFC to the CNVR 326 and receives the signal OVR from the temperature sensor in the radio unit 206. The Tx control 338 provides control signals for the transmitter 220. The PLL/DDS control 340 provides . the control signals to the synthesizer 222 to control the generation of the synthesized frequencies. The DSP bus interface 344 provides an interface between the ASIC 328 "and the digital signal processor 330. The Viterbi decoder 346 is utilized to process signals. The timing & interrupt controller 348 provides timing signals.

Fig. 20 illustrates the BBU controller 228. The controller 228 includes a microcontroller circuit 352, a MMI interface circuit 354, a car accessories interface circuit 356, a microprocessor supervisor 358, a power supply control circuit 360 and a memory circuit 362. The microcontroller 352 is preferably implemented with a Motorola processor MC68302 or an Intel 386EX. The memory 362 is preferably comprised of 128k x 8 static RAM and 512k x 8 flash memory. The memory 362 holds the programs of the microcontroller 352, of the modem 226 and of the VPP 230. The memory 362 will also hold various parameters, user defined telephone numbers and messages that should be kept nonvolatile.

The power supply control circuit 360 monitors the state of the subscriber unit 12 and the state of the car and controls the power supply. The supervisor circuit 358 is responsible for the reset mechanism and the power fail indication. The microcontroller 352 is connected to the MMI interface 354 via a single ON/OFF input and via two RS-232 lines, one for each direction. The RS-232 lines carry data, control and test messages. The car accessories interface circuit 356 provides one input from the car's ignition switch and one output to the car's horn. The microcontroller 352 has a two-wire bi-directional RS-232 connection available for GPS interface.

Referring to Fig. 21, a block diagram of the voice processing package (VPP) 230 is illustrated. The VPP 230 performs encoding and decoding of voice signals and consists of a CODEC 364, an analog interface circuit 366 and a digital signal processor 368. The digital signal processor 368 is preferably an Analog Devices' ADSP2115. The program for the digital signal processor 368 is stored in the memory 362 of the BBU controller 228.

Reference is now made to Figs. 22A and 22B which illustrate a preferred service board found in the subscriber unit.

Figs. 23 and 24 show, respectively, the preferred signal processing performed when transmitting and receiving signals on traffic channels.

Signal Processing Figs. 25 to 31 illustrate various coding schemes utilized in the present invention. Fig. 25 shows the preferred transmit processing steps for a voice signal when the system is operating in the differential mode. After the signal is voice encoded in step 550, the bits of the signal to be transmitted are divided into classes in accordance with their importance. Those bits with greater importance then receive greater coding. In accordance with a preferred embodiment, four classes (I, II, III and IV) are utilized. The twelve most important bits are as- signed to Class I. The next 36 most important bits are assigned to Class II. The next 32 most important bits are assigned to Class III. The least 8 important bits are assigned to Class IV. The bits are then coded with a CRC encoder in step 552 and with a convolutional encoder in step 554 as illustrated. Inband control signals received are also encoded as illustrated. Then, in step 556 the signal is further coded with a partial repetition encoder by the repetition of 10 symbols. In step 558, a symbol is replaced with the value P R_CNT_SYS with a puncture encoder. The symbol that is replaced is preferably one of the lesser important bits. In step 560 a permutation step can be performed, however, it is presently preferred not to perform this step. In step 562, the interleaving already discussed is performed. In step 564 the frame of data is converted into time slots for transmission.

Fig. 26 illustrates the steps performed by the Rx processing unit in the differential mode. The steps are essentially the opposite of those described with respect to Fig. 25. In step 566, the slot of data is properly framed for processing. In step 568, the data is deinterleaved using the reverse process used during the transmit processing. In step 570, de-permutation is performed if the permutation step was performed during the transmit processing. In step 572, the symbol PWR_CNT_SYM is removed. Then in steps 574 to 576, the signal is decoded as illustrated. In step 578, the bit error rate is calculated and, if too high, can disable the voice reception .

Figs. 27 and 28 illustrate the steps performed by the Tx processing unit and the Rx processing unit, respectively, during the coherent mode of operation. These steps are very similar to those already discussed with respect to Figs. 25 and 26, except the partial repetition encoder repeats a different number of bits during its encoding.

Figs. 29 and 30 illustrate the preferred steps performed when processing data in the differential and in the coherent modes, respectively. Referring to Fig. 29, the data contains a header and one or more frames of data per message.

The data is subdivided into a number of frames, each of which is CRC encoded as illustrated. The header is also CRC encoded, as shown. The remaining encoding steps are selectable, the selection depending on the channel conditions. In the low rate mode, the remaining coding includes convolutional encoding (with r = 1/4) and repetition encoding (rate equal 1/2) , prior to interleaving, as illustrated. In the medium rate mode, the remaining coding includes convolutional encoding (with r = 1/4) prior to interleaving. In the high rate mode, the coding includes convolutional encoding (with r = 1/2) prior to interleaving. In the very high rate mode, the coding rate is one. The coding for the coherent mode is very similar to the coding utilized in the differential mode, as illustrated in Fig. 30.

The preferred coding scheme utilized for the ACH and the CCH is illustrated in Fig. 31.

Other Processes Correlations Performed on Received Data The subscriber unit 12 performs several correlations on the signals it receives in order to achieve proper synchronization with the signal transmitted by the base station 10. In accordance with a preferred embodiment of the present invention, the complex conjugates of the synchronization signal are formed as follows: HO = -1 + j ; HI = 1-j ; H2 = -1-j ; H3 = 1-j ; H4 = 1+j ; H5 = 1-j ; H6 = 1+j; H7 = -1-j; H8 = 1+j; H9 = -1+j; H10 = 1-j; Hll = 1+j; H12 = -1+j; H13 = 1-j; H14 = -1-j; H15 = -1+j; H16 = -1-j ; H17 = -1+j and H18 = -1-j.

The incoming signals are correlated according to the following formulas: RL_COR0 = The sum of Re{U0 (p-j*4) } *Re{Hj } - Im{U0(p-j*4) )*Im{Hj } , where the sum limits are from j = 0 to j 18; RL_C0R1 = The sum of Re{Ul (p-j*4) ) *Re(Hj ) - Im{Ul(p-j*4) }*Im{Hj ) , where the sum limits are from j = 0 to j 18; where p = 82, the index of the last sample of the synchronization code, UO(n) is the data from the first receiver and Ul(n) is the data from the second receiver.

Next, diversity selection is performed, if RL_COR0 is greater than RL-C0R1, then RL_COR_PEAK = RL_COR0 and set Ui = UOi for i = 4 to 163, otherwise RL_COR_PEAK = RL_C0R1 and set Ui = Uli for i = 4 to 163.

Then the following correlations are performed: RL_COR_EARLY = RL_COR(p + 1) RL_COR_LATE = RL_COR(p - 1) by using Ui with i = 4 to 163.

Note that these correlations are offset in time from the earlier correlations described.

In addition, the following correlation is performed: IM_COR-PEAK = The sum of Re {U (p-j ) } *Im{ Hj } +Im{ U (p-j *4 ) } *Re {Hj } , where the sum limits are from j = 0 to j = 18.

Channel Acquisition One of the processes which the subscriber unit 12 must perform prior to starting communications is the acquisition of the timing of the base station 10. This channel acquisition process is performed when the power to a subscriber unit 12 is first turned on and any time synchronization is lost.

Generally, a subscriber unit 12 performs acquisition by first accessing a table of communication sites which is maintained in the memory of the subscriber unit controller 228. This table contains a list of sites and the frequency of the associated CCHs. The subscriber unit 12 then sequentially scans the frequencies of channels identified by the site table, trying to lock onto each one in turn. In particular, the subscriber unit 12 attempts to lock onto a channel by searching for the synchronization pattern in a SLS slot in the channel. When the subscriber unit 12 scans a frequency, if there is no CCH at that frequency, there will be no SLS to detect. However, when the subscriber unit 12 scans a control channel (CCH) it will find a SLS. Generally, the subscriber unit 12 performs the scanning by performing a sliding correlation based on the mean square of the phase error between the signals on the channel being scanned and the known sync pattern of the SLS slot which the subscriber unit 12 has stored in its memory. When the subscriber unit 12 finds a high correlation, it has found the SLS slot. The subscriber unit 12 can also search all possible frequencies if the search of known sites is not successful.

Referring to Fig. 32, the process of channel acquisition is illustrated. In step 600, a counter, SLOT_COUNT, is initialized. In step 602, the signals (W1(P)) and the diversity signal (W2(P)) being received by the receivers 214 and 216 are first filtered with a digital matched filter and then differentially demodulated. The filter outputs, Zl(p) and Z2(p) and the demodulation outputs Ul(p) and U2 (p) are output to step 604. In step 604, the average amplitude of these outputs from each receiver channel are compared and the output with the highest average amplitude is selected.

In step 606, the sliding correlation is performed on the data from the selected channel. The correlation measures the mean square phase error between the phase of the received signals and the phase of the synchronization code which is stored in memory.

In step 608, a search is performed for the optimum correlation by looking for the minimum error. The process searches for several minimums in a 216 slot window. Generally, approximately five minimums will be found in this window.

Then, in step 610, the process determines whether the prior steps have been performed the prerequisite number of times. If they have not, then the steps are repeated. If they have, then the process goes on to step 612. Here, the process verifies that it has locked onto the proper synchronization pattern in the SLS slot. The time index of the minimums in the frame of data has been stored. The time between two minimum values must have a legitimate value (generally 72 slots) to be verified.

In step 614, if it is determined that proper synchronization has not been obtained, the process informs the controller 228 that synchronization has failed. On the other hand, if in step 614, it is determined that proper synchronization has been obtained, then the process continues to step 616. In step 616, the timing of the slot is adjusted in accordance with the indices of the minimum received value. In other words, the synchronization timing is known and the slot timing is determined from this knowledge.

Delay Lock Loop To lock onto the synchronization information in the SLS slot of the CCH, the subscriber unit 12 performs a delay lock loop process. Referring to Fig. 33, a preferred delay lock loop process for the subscriber unit 12 is illustrated. The correlation values, RL_COR_EARLY and RL_COR_LATE, are utilized in step 618. The following calculation is performed to determine the delay: (RL_COR_EARLY - RL_COR_LATE) / (RL_COR_EARLY + RL_COR_LATE) .

When the synchronization information is being correlated, this calculation will yield a result that is nearly zero. On the other hand, if synchronization is not being correlated, then the result will be non-zero.

The result of this calculation is filtered with an infinite impulse response filter in step 620 to provide a smooth response. Then, in step 622, the filtered result is input to a number controlled delay generator where a controlled delay is introduced.

Referring to Fig. 34, a preferred delay lock loop process for the base station 10 is illustrated. Essentially, the same steps 618, 620 and 622 are implemented. However, the base station 10 implements these steps in software so an additional integrating step, step 624, is required.

Automatic Frequency Control The subscriber unit 12 performs an automatic frequency control process to measure and correct the frequency inaccuracies of its frequency source. Referring to Fig. 35, the steps performed by the subscriber unit 12 during the differential mode are illustrated.

The correlations RL_COR_PEAK and IM_COR_PEAK, of the received data are utilized to determine the errors in the frequency being received by the subscriber unit 12. The error is determined, in step 626, by inputting these values into a table to determine the arctangent of the phase difference between RL_COR_PEAK and IM_COR_PEAK. Then the arctangent is filtered in step 628, as illustrated. The result is utilized to correct the frequency source at the subscriber unit 12.

Initially, it is preferred to estimate and correct the frequency error. This estimation is also utilized as an initial condition at which to start the previously described frequency control process. The first step in the estimation is to select the complex samples of the currently active time slot.

The frequency of the slot and nominal value of the slot frequency are determined and, via software, the several frequency values which vary around the nominal value are substituted into the slot. For example, the nominal frequency and six frequencies above the nominal may be selected. The mean square phase error of each of these frequencies is calculated and the frequency with the minimum value is utilized.

Automatic Gain Control The receivers in the base station 10 and the subscriber unit 12 all need to receive signals within a certain amplitude range in order to be able to optimally perform demodulation. Gain control, therefore, is performed in both the base station 10 and in the subscriber unit 12 to maintain a constant signal level .

Referring to Fig. 36, the gain control process for a receiver channel in the subscriber unit 12 is illustrated. In the first step 630, the average amplitude of each time slot is calculated from the I and Q inputs associated with each symbol. Specifically, this is accomplished by summing the value r(n) = l(n) + jQ(n) ror each time slot n and then dividing by the number of symbols in a time slot, in this case, 39.

Then, in step 632, the average amplitude is compared to a reference level. It is preferred to set the reference level at 0.25, which is 12 dB below the maximum level of 1. The difference between the average amplitude and the reference level is filtered in the next step 634 to smooth the gain control operation to operate on slower variations. The filter is preferably implemented as a standard infinite impulse response filter with a gain of 10 and the coefficients of = 0.95 and C2 = 1 - CQ_ = 0.05. In addition to the filtering, in step 636, it is preferred to limit the range of the signal being filtered to the range of the input signal. In Fig. 36, the signal is limited to the range of +0.5 to -0.5.

The output of the filter is then sent to a look up table in step 638. The look up table is used to account for the nonlinearities found in the amplifiers in the receivers. Accordingly, the values of the look up table will vary according to the amplifiers being used in the receivers. Then the output of the look up table is converted to an analog signal which is, in turn, applied to the amplifiers in the receivers to effect gain control of the received signal.

In the subscriber unit 12, the steps 630 to 636 are performed in the digital signal processor (DSP, Fig. 15) of the modem 226. The value from the filtering step is output to the receiver channels 214 and 216 (Fig. 16) in the subscriber unit 12 through the ASIC 328 (Fig. 18) and through the CNVR INTERFACE 336 (Fig. 19) .

It is preferred to utilize the calculation obtained from a time slot to control the gain of the next slot. Also, the processing is performed separately for each channel, that is, one calculation is performed on the signals received from the receiver channel 214 and another calculation is performed on the signals received from the diversity channel 216.

The previously described steps are performed once the gain control has been performed. During this time, the path to the look up table AGC_TBL is through the limiter 636. During power up, however, the path to the look up table AGC_TBL is through the line labeled SGCS_IC and it is preferred to perform a different gain control process.

Referring to FIG. 37, the preferred steps of the initial gain control process for the subscriber unit 12 are illustrated. These steps set the initial value of SGCS_IC and of the memory Z-1 xn FIG. 36. In step 640, the initial value of SGCS_IC is set to the middle of the dynamic range of the input signal, in this case 0. Then, in step 642, the SGCS_IC is applied to the AGC_IC look up table, as illustrated in FIG. 36.

Next, in step 644, a sliding window, preferably of 26 msec. in duration, is established. During a first part of the window, the average value of 164 symbols is determined. Then in the next part of the window, the average value of the next 164 symbols is determined. In a preferred embodiment, this step.: is repeated a total of 1476 times. The maximum of each of these averages, MAX_W is determined.

Then in step 646, the maximum value from step 644, MAX_W is compared to the predetermined value Wl, which is preferably 0.6 and to the predetermined value W2 , which is 0.1. If MAX_W is greater than Wl, then in step 648, the initial value of SGCS_IC is increased by a . In a preferred embodiment , = 0.09. Then, in step 650, the value of SGCS_IC is checked. If the value is less than or e.qual to 0.5, then the process returns to step 642. However, if the value is greater than 0.5, then step 652 is performed. In this step, SGCS_IC is set equal to 0.5, the maximum value of the dynamic range of the signal. Then in step 654, the memory Z-1 (FIG. 36) is set equal to 0.5 and the signal SGCS_IC is applied to the look up table AGC_TBL.

If in step 646, MAX_W is determined to be less then W0, then in step 646, the initial value of SGCS_IC is decreased by a , which is preferably 0.09. Then, in step 648, the value of SGCS_IC is checked. If the value is greater than or equal to -0.5, then the process returns to step 642. However, if the value is less than -0.5, then step 650 is performed. In this step, SGCS_IC is set equal to -0.5, the minimum value of the dynamic range of the signal. Then in step 654, the memory Z-1 (FIG. 36) is set equal to -0.5 and the signal SGCS_IC is applied to the look up table AGC_TBL.

If in step 646, MAX_ is determined to be less than or equal to 0.6 or greater than or equal to 0.09, then in step 652, the value of SGCS_IC is incremented by a value which is a function of the value of MAX_ . Then in step 654, if SGCS_IC is greater than 0.5, the process goes to step 652. On the other hand, if SGCS_IC is less than or equal to 0.5 in step 654, then the process goes to step 656. In step 656, if SGCS_IC is less than -0.5, the process goes to step 650. On the other hand, if SGCS_IC is greater than or equal to -0.5, then step 654 is performed. In step 654, the memory Z-1 (FIG. 36) is set equal to SGCS_IC and the signal SGCS_IC is applied to the look up table AGC_TBL.

The same basic steps are utilized to perform gain control in the base station 10. There are, however, some differences in the implementation details. For example, the averaging step and look up table are implemented in the slot processor while the remaining steps are implemented in the frame processor. Also the initial condition is set via a message on the ACH which set the receiver at the base station 10 at a fixed value.

Hand Off The communication system of the present invention must account for the movement of subscriber units 12 between the sectors 14 to 16. The process by which the movement is accounted for is referred to as the "hand off". It is preferred that the hand off process be seamless, that is, those persons using the system to communicate should not be affected by this process.

Generally, the hand off process includes first detecting when a situation requiring hand off occurs. Simply stated, hand off situations occur in the sectorized system of the present invention when a subscriber unit 12 relocates from one sector to another, say from sector 14 to sector 15 (FIG. 1) or between any type of site or sector. The next step in the hand off process is to then perform the necessary switching to allow seamless hand off of the communications between the sectors.

In the present invention, the hand off situations are detected by the subscriber unit 12. The subscriber unit 12, by virtue of the channel acquisition procedure and the delay lock loop tracking procedure, is already detecting and locked onto the SLS slots of the CCH in the sector in which it is located, say for example, sector 14. The subscriber unit 12 also searches for the SLS slots of the CCH in the sectors which are adjacent to the sector in which it is located, that is the SLS slots of the CCH in sectors 15 and 16 are searched for. The subscriber unit 12 accomplishes this by referring to the label tag embedded in each SLS and determining the frequencies of the CCHs in the adjacent sectors 15 and 16 of the site 1 or, if the CCHs are all on the same frequencies, by looking at that frequency.

The subscriber unit 12 will, therefore, have three data streams to process — i.e., the SLS data from the sector in which it is located (sector 14) , the data from a first adjacent sector (sector 15) and the data from the second adjacent sector (sector 16) . The digital signal processor 226 in the subscriber unit 12 performs an average correlation of synchronization from each of these data streams as compared to the known synchronization pattern to detect when a hand off situation is occurring. The average correlation of the information from the sector 14 is obtained from the correlation processes previously described.

The correlation of the information from the sectors 15 and 16 are somewhat different because the timing of these channels is not known with certainty. Several correlations are performed on the synchronization information from these sectors based on a prediction of the timing of these synchronizations. Then, for each sector, the correlation with the maximum value is selected. Then, the maximum value for several correlations within that sector are selected and averaged.

While the subscriber unit 12 is in the middle of a sector 14, the average correlation of the synchronization infor- mation from that sector will be higher than the average correlation of the synchronization information from the adjacent sectors 15 or 16. However, as the subscriber unit 12 approaches a new sector, say sector 15, the average correlation of the synchronization information from the new sector will be increasing. Eventually, the average correlation of the synchronization information from the new sector 15 will exceed the average correlation of the synchronization information from the old sector 14. When the average correlation of the information from the new sector 15 exceeds the average correlation of information from the old sector 14 by a predetermined amount, the subscriber unit 12 determines that a hand off situation exists.

Once the subscriber unit 12 determines that the hand off situation exists, the controller 228 causes a message to be transmitted on the ACH to the base station 10. The message requests the initiation of a hand off procedure by the base station 10. The base station 10 acknowledges the request by the subscriber unit 10 on the CCH.

Once the base station 10 receives the request of the subscriber unit 12 to initiate a hand off, the base station 10 selects a communication link in the new sector 15 and connects it with the communication line in the old sector 14 via a three way conference bridge found in the PABX 36. In this way, a communication link is established between the subscriber unit 12 in the old sector 14, the subscriber unit 12 in the new sector 15 and the other person (or machine) communicating with the subscriber unit 12, thereby enabling a seamless hand off.

At this point, the base station 10 and the subscriber unit 12 are both ready to accomplish the hand off. In accordance with the present invention, the hand off is accomplished during a time when there is no activity on the channel. This time is detected by the voice activity detector (VAD) in the vocoder in both the base station 10 and the subscriber unit 12. It should also be noted that the hand off is done independently in the uplink channel and in the downlink channel. Therefore, the VAD in the base station 10 detects the voice inactivity in the down link communications and, upon detection, causes the hand off to occur between the downlink channels. Also, the VAD in the subscriber unit 12 detects the voice inactivity in the uplink communication and upon that detection, causes the hand off to occur between the uplink channels.

The actual process by which switching occurs varies. In both the base station 10 and in the subscriber unit 12, the VAD provides a bit upon detecting voice inactivity to the transmitters. The transmitters, upon sensing the bit from the VAD, switch the transmission from the old sector to the new sector. The receiver hand off is accomplished by transmitting a STOP RECEIVE marker in the TCH which tells the receiver that a talk spurt has ended — i.e. that there is voice inactivity — so that it is time to switch to the new sector.

In the event that the receiver misses the marker on the TCH, the hand off can be accomplished by another technique. In accordance with this alternate technique, the communications being received by the receiver includes a CRC code, as previously discussed. This CRC code is analyzed to determine whether the communications has been properly received. For the hand off process, it is determined how many CRC failures have been detected. When the number of CRC failures exceeds a certain amount during the hand off process (the subscriber unit 12 has requested a hand off) , then the hand off is automatically accomplished.

Power Control The base station 10 and the subscriber unit 12 both perform power control processes whereby each communication link transmits at a minimum power needed for transmission, thereby minimizing the interference in the communication system and saving power. It is preferred to provide power control for the TCH and the ACH, but not for the CCH.

The power control for the downlink TCHs will now be described with reference to FIG. 38. The downlink TCH transmissions (from the base station 10 to the subscriber units 12) start at predetermined power, preferably at the maximum power. In step 660 of the power control process, the subscriber unit 12 detects the received power. In a preferred embodiment, the average of the power values (SGCSO and SGCS1) determined for the two receiver channels 214 and 216 during the automatic gain process are utilized in this process.

Then, in step 662, the average power from step 660 is compared to a predetermined threshold, REF, and the difference is filtered in step 664 so that power is controlled in a smooth fashion. The output of the filter, FIL_SGCS, is then sent to the transmitter in the base station 10 to modify the transmitted power. If there is no activity on the TCH, the information is sent to base station 10 via the ACH. In this case, it is preferred to accumulate the differences between the sensed power and the predetermined threshold so as to modify the transmitted power in increments, for example by 5 dB. If, however, the TCH is active, then the information is preferably sent via the TCH. In this case, it is preferred that the transmitted information be controlled to modify the transmitted power in small increments, for example by 1 dB.

It is preferred to utilize a different method to control the power transmitted on the ACH due to short message size available on the ACH. The subscriber unit 12 first measures the average received power of the signal being transmitted on the CCH from the base station 10. The CCH transmission power, which is constant, is broadcast and received by the subscriber unit 12. The subscriber unit 12 then subtracts the known CCH transmission power from the average received power to determine the transmission losses on the CCH. The desired reception power at the base station 10 of signals transmitted by the subscriber unit 12 is known by virtue of a broadcast message. The subscriber unit 12 adds the determined transmission losses on the CCH to the desired reception power to determine the power at which the subscriber unit 12 will transmit on the ACH.

The control of the power transmission on the uplink TCH transmissions (subscriber unit 12 to base station 10) is preferably controlled using a combination of the previously described methods. When the subscriber unit 12 first starts transmitting to the base station 10 on the uplink TCH, the power is controlled in the same way as power is controlled on the ACH. After this starting point, the power on the uplink TCH transmissions is controlled using the same process used to control the power on the downlink TCH transmissions.

Reference is now made to Figs. 39 - 48B, which are simplified block diagrams illustrating the structure of the software components of the system of Fig. 1.

The structure and operation of various elements of Figs. 39 - 48 are described in the following appendices: Element 704 of Fig. 39 is described in Appendix 2. 1. 30.

Element 710 of Fig. 41 is described in Appendix 2. 1. 32.

Element 712 of Fig. 41 is described in Appendix 2. 1. 31.

Element 722 of Fig. 42 is described in Appendix 2. 1. 05.

Element 724 of Fig. 42 is described in Appendix 2. 1. 11.

Element 728 of Fig. 43 is described in Appendix 2. 1. 01.

Element 730 of Fig. 43 is described in Appendix 2. 1. 15.

Element 732 of Fig. 43 is described in Appendix 2. 1. 08.

Element 734 of Fig. 43 is described in Appendix 2. 1. 20.

Element 736 of Fig. 43 is described in Appendix 2. 1. 07.

Element 738 of Fig. 44 is described in Appendix 2. 1. 02.

Element 740 of Fig. 44 is described in Appendix 2. 1. 03.

Element 742 of Fig. 45 is described in Appendix 2. 1. 12.

Element 744 of Fig. 45 is described in Appendix 2. 1. 06.

Element 748 of Fig . 46 is described in Appendix 2. 1. 17.

Element 750 of Fig. 46 is described in Appendix 2. 1. 18.

Element 752 of Fig. 46 is described in Appendix 2. 1. 19.

Element 754 of Fig . 46 is described in Appendix 2. 1. 04.

Element 758 of Fig. 46 is described in Appendix 2. 1. 16.

Element 760 of Fig. 47 is described in Appendix 2. 1. 09.

Element 762 of Fig. 47 is described in Appendix 2. 1. 10.

Element 764 of Fig. 47 is described in Appendix 2. 1. 13.

Element 766 of Fig. 48 is described in Appendix 2. 1. 34.

Element 768 of Fig. 48 is described in Appendix 2. 1. 35.

Referring ;again to F.igs . . 8 - 9, various elements Of said Figs, are more fully described as follows: Elements 56, 57 and 58 of Fig. 9 may be any suitable UHF Antenna, such as model DB 874H105, commercially available from Decibel Products, 3184 Quebec Street, POB 569610, Dallas, TX, USA.

Element 38 of Fig. 8 may be any suitable voice mail system, such as a Trilog system, commercially available from Comverse, Israel.

Element 36 of Fig. 8 may be any suitable PABX system, such as a Coral III or Coral IV system, commercially available from Tadiran, Israel.

Reference is now made to Figs. 49 - 51 which are simplified block diagrams illustrating the structure of a microsite .

Unless otherwise specified, microwave antennas in the apparatus of the present.: application may be any suitable microwave antenna, such as P/N 8838A-24/PCN, commercially available from Alpha Industries, Inc. , Massachusetts, USA.

The apparatus of Figs. 49 - 51 comprises a UHF antenna, which may be any suitable UHF antenna, such as model DB808-Y, commercially available from Decibel Products, 3184 Quebec Street, POB 569610, Dallas, TX, USA.

Conventional techniques for extracting side information (channel state information) include the following publications, all of which are hereby incorporated by reference: [1] Viljo Hentimen, "A Channel Feedback Communication System", ACTA POLYTECHNICA SCANDINAVICA , Electrical Engineering series No. 26, Helsinki 1971. [2] A.J. Viterbi, "A Robust Ratio-Threshold Technique to Mitigate Tone and Partial Band Jamming in Coded MFSK System". MILCOM 82, Boston MA, October 18-20, 1982. [3] Hyuck M. Kwon, "Imperfect Jamming State Information Generators for Coded FH/MFSK under Partial-Band Noise Jamming". MILCOM 89, pp. 1.1.1 - 1.1.5. [4] Hyuck M. Kwon, "Capacity and Cutoff Rate of Coded FH/MFSK Communications with Imperfect Side Information Generators", IEEE Journal on Selected Areas in Comm. Vol. 8, No. 5, June 1990, pp 750-761. [5] Hyuck M. Kwon, Pil Joong Lee, "Combined Tone and Noise Jamming Against Coded FH/MFS ECCM Radios", IEEE Journal on Selected Areas in Comm. Vol. 8, No. 5, June 1990, pp. 871-883. [6] R.R. McKerracher, P.H. Wittke, "Frequency Hopped Spread Spectrum Systems with Reed Solomon Coding and Practical Jammer State Estimation", MILCOM 92, pp. 16.6.1 - 16.6.6. [7] G.L. Stuber, Jon W. Mark, Ian F. Blake, "Diversity and Coding for FH/MFSK Systems with Fading and Jamming - Part II: Selection Diversity", IEEE Trans, on Comm., Vol. 37, Aug. 89, pp 859-869. [8] Joseph M. Hanratty, Gordon L. Stuber, "Performance Analysis of Hybrid ARQ Protocols in a Slotted Direct Sequence Code-Division Multiple Access Network Jamming Analysis", IEEE Journal on Selected Areas in Comm., Vol.8, No. 4, May 1990.

Reference is now made to Fig. 52, which is a simplified block diagram illustrating the structure of a remote sector. A remote sector, also termed herein a "remote station", functions to enhance the coverage of the system. The apparatus of Fig. 52 is described in Appendix RS .

Time Alignment Method In accordance with a preferred embodiment of the present invention, the system is operative to straighten out the timing of all of the signals received by the base station.

Generally, subscribers are at different distances. They all downlink and are synchronized. They send uplink on the synchronized timing but due to different distances the communications are not received in synchronized timing.

A preferred method to make sure that they are received with synchronize timing is as follows: The base receiver measures the time of arrival on the uplink, and checks if they are within a predefined window. The base station tells the remotes to change their transmission timing so as to have the received signals within the window of the synchronized timing of the base station.

The above method is described more fully in the Algorithms Document appended hereto.

MARO Method In accordance with a preferred embodiment of the present invention, the system is operative to reduce the need for large number of multiple retransmissions when long digital messages are being sent.

A preferred method for reducing the large number of retransmissions is as follows: Break received message into segments and flag only those segments which are not received correctly. Upon each subsequent transmission only replace those flagged portions; once all flagged portions are received correctly, there is no need for further retransmission even though the retransmission may not have been received totally completely.

What is claimed is:

Claims (9)

1. S. 111340/ 1 is claimed is: 1 . A method of controlling the transmi tted power in a communication system, comprising the staps of: transmitting a signal at a first power from a first transmitter; receiving the signal and determining The pcwar of the signal received; comparing the power of the signal received to a predetermined threshold and sanding the difference signal to the first transmitter: changing the power of the signal transmitted from the first transmitter on an incremental basis.

2. The method of claim 1 , wherein the first power is the maximum power.

3. A method of controlling the power of a signal transmitted from a first communication station to a second communication station wher9 the power of the signal transmitted from the second communication station to the first communication station is known by the first communication station, comprising the steps of: when the second communication station transmits to the first com munication station, determining the power cf the signal racaived at the first ccmm unicaticn station; 59 111340/1 comparing the power of the signal transmitted by the second ccmmunication station to the power receivad by the first communication station to determine the transmission losses; adding the transmission losses cf the signal transmitted by the second ccmmunication station to the first communication station to the desired reception power of a signal transmitted by the first communication station to the second communication station to determine the transmission power of the first ccmmunication station.

4. , The method of claim 3, further comprising the steps of: after a predetermined amount of time, transmitting a signal at a first pcwar from a first transmitter; receiving the signal and detarmining the power of the signal received; comparing the power of the signal received to a predetermined threshold and sending the difference signal to tha first transmitter; changing tha power of tha signal transmitted from the firet transmitter on an incremental basis.

5. A method of handing off communications between a communicating party and a mobile communication radio which is mcving from a first sector to an 60 111340/1 acjacent sector in a communication base s :a:ion having multiple sectors, cinn prising the steps of : the mobile radio detecting synchro nization in formation from the first sector and searching for synchronization information from sectors adjacent the first sector; the mobile radio determining when the synchronization information from an adjacent sector has stronger reception than the synchronization information from the firs sector and then requesting a hand off from the base station; the base station enabling a three way communication link between the mcbile radio in the first sector, the radio in the adjacent sector and the communicating party; monitoring voice activity from the base station to the mobile radio and handing off downlink communications rom ;ha base station to the mcbile radio from ne first sector to the adjacent sector when voice inactivity is detected; monitoring voice activity from the mobile radio to the base station and nanding off uplink communications from the mociie radio to the base station from the first sector to the adjacant sac cr when voice inactivity is detected. c .

6. A method of controlling the gain of a time sic* in a time multiplexed ic mmunication system, comprising the steps of : averaging the amplitude values c symbols associated with a racsived ;~ 5 slot; 61 111340/2 comparing the average ampli tud e wi th a ref erence signal to obtain the d^erence; filtering tha difference signals; adjusting the gain of a following time slot.

7. The method of claim 2, further comprising the step of: limiting the range of the signal being filtered.

8. The method of claim 1 , wherein the reference signal is 1 2 dB below the maximum value of the input signal.

9. The method of claim 1 wherein the gain of the time slot immediataly following the received time slot is controlled. 1 0. The method of ciaim 1 , further comprising the step of adjusting the output cf the filtered signals to account for ncnlinearitias in the amplifiers. For the Applicant, & Co. C: 20385 62