GB2101851A

GB2101851A - Bit synchronization adjuster

Info

Publication number: GB2101851A
Application number: GB08118501A
Authority: GB
Inventors: Graham Edgar Beesley
Original assignee: Motorola Ltd
Current assignee: Motorola Solutions UK Ltd
Priority date: 1981-06-16
Filing date: 1981-06-16
Publication date: 1983-01-19
Also published as: EP0081500A1; WO1982004515A1; JPS58501488A

Abstract

A data processing system of the kind in which there is a transmission of a code format having a pattern that provides a plurality of reversals 1011010 etc. from which it is desirable to maintain bit synchronization for the incoming data once initial synchronization is obtained. The system being characterized in this that after a first synchronization codeword is recognized at least a part of any following synchronization codeword is sampled at least twice (or three times) per bit and compared with the known structure of the bits of a known synchronization codeword to enable the bit structure of best fit with the known bit structure to be accepted by the system to sample the incoming data. The best fit may conveniently be a running total or summation of least errors when compared with the structure.

Description

SPECIFICATION Bit synchronization adjuster This invention relates to a data processing system of the kind (hereinafter referred to as the kind set forth) in which there is a transmission of data of a code format having a pattern that provides a plurality of reversals 1011010 etc. from which it is desirable to maintain bit synchronization for the incoming data once initial synchronization is obtained.

The desideratum is achieved according to the present invention by a data processing system of the kind set forth wherein after a first synchronization codeword is recognized at least a part of any following synchronization codeward is sampled at least twice per bit and compared with the known structure of the bits of known synchronization codeword to enable the bit structure of best fit with the known bit structure to be accepted by the system to derive the best sample instant of the transmission of said data.

The best fit may conveniently be a running total or summation of least errors when compared with the structure.

In a preferred embodiment the codewords are sampled not twice but three times of which one time is both early and late of the other times and comparison is performed with the known structure of the synchronization codeword.

Ideally the said one time is taken substantially at the position corresponding to the optimum sampling time obtained from the previous synchronization.

One satisfactory analysis of a 32 bit synchronization codeword, for example, is to take three samples, of which the inner is both early and late of the two other samples by say 1/8th of a bit width. Other spacings than 1/8th may be used to advantage but owing to drift error it should not come too close to the value of the said error. For example where the drift error is about 1/13th of a bit width, the sample spacing should exceed that value.

The invention will be more fully understood from the following description given by way of example only with reference to the figures of the accompanying drawing in which Figure 1 shows seven bits of a 32 bit synchronization codeword.

Figure 2 is a functional block diagram of one bit synchronization adjuster of the invention and Figure 3 is a flow diagram of operations to be performed by the memory and microprocessor of Figure 2.

The substantially centrally placed sample is conveniently called the inner sample (I) and the sample early of I (E) and the sample late of I (L). Error computation is given with relation to Figure 1 of the drawing and this is compared with the known bitstrucureofthesynchronization codeword which is shown as K. It is readily seen from the figure that the received synchronization codeword R has blemishes due to noise and is sampled three times at I.E. and Where I is substantially centrally placed of the bit width Bathe distances between I and E and I and L being in each case 1/8 Bw.

The error of the sampled structure from the known structure K is taken to be unity e1. Where the error is non-existent and the structure correct as in Kthe error is zero, eO.

The following table is then readily constructed: B1 B2 B2 B4 B5 B6 B6 Total Errors E e0 e0 e1 e1 e0 e1 e0 3 e1 e0 e0 e1 e0 e1 e0 3 L e0 e0 e1 e0 e0 e1 e0 2 Clearly the sample or version at L has the least errors in comparison with the bit structure of K and thus L is the best fit and gives the optimum sampling instant and is subsequently accepted for use in the system.

The circuitry of one data processing system operates as follows: During the reception of the codeword a number of separate sampling instants are used and repeated in the same relationship for each bit of the codeword. In this way a number of versions of the received codeword may be built-up one corresponding to each of the sampling instants used. The version of the codeword having the least errors and giving the best fit with the known structure of the codeword will then be accepted by the system to indicate the best sampling instant. This instant and its repeat once every bit is then used to sample the incoming data.

Once specific embodiment of this function utilizes a microprocessor based system (Figure 2). The flow chart associated with this function is Figure 3 and a listing is attached in the appendix A.

In Figure 2 a radio pager comprises a receiver board 10 having an aerial 11 by which a signal enters a radio frequency amplifier 12 a 1st mixer 13 coupled to a 1st oscillator 14to a 1st intermediate frequency amplifier 15; from which it passes inter alia via a 2nd mixer 16 to an audio limiter 17 connected to the input/output of the microprocessor 18 (MC 146805E2) which is related to the memory of ROM (MCM65516) 19.The MPU will receive from the ROM a logical sequence set out in the flow sheet of Figure 2 wherein 101 is input first bit initialize error registers 102 is retain LSB invert initialize bit count 103 is advance timer 1/8th of a bit 104 is initialize X for three samples 105 is wait 106 is input data 107 is EOR with expected bit 108 is add result to error register 109 is three samples in? 110 is thirty two bits in? 111 is initialize for early (E) sample selection 112 is does late (L) sample have less errors? 113 is SET accumulator and index register for late (L) timing 114 is does late (L) sample have same errors? 115 is set accumulator and index register for innertiming (I) 116 is does inner (I) sample have less errors? 117 is put next expected bit into W2SWB 118 is reset dead man timer 119 is minimum error count less than W2EROR? 120 is adjust timer according to accumulator contents 121 is set for BFLAG 122 is RETURN

Claims

1. A data processing system of the kind set forth wherein after a first synchronization codeword is recognized at least a part of any following synchronization codeword is sampled at least twice per bit and compared with the known structure of the bits of the known synchronization codeword to enable the bit structure of best fit with the known bit structure to be accepted by the system to derive the best sample instant of the transmission of said data.

2. The data processing system as claimed in claim 1 wherein the codeword is sampled not twice but three times of which one time is both early and late of the other times and comparison is performed with the known structure of the synchronization codeword.

3. The data processing system as claimed in claim 2 wherein the said one time is taken substantially at the position corresponding to the optimum sampling time obtained from the previous synchronization.

4. The data processing system as claimed in claim 2 or claim 3 wherein the samples are taken at a separation of 1.8th of a bit width.

5. The data processing system as claimed in any preceding claim wherein the said best fit is a running total or summation of least errors.

6. A data processing system constructed and arranged substantially as hereinbefore described and as shown in the figures of the accompanying drawing and the appendix.