• No results found

Description of a simple experiment of a TDMA synchronisation principle

N/A
N/A
Protected

Academic year: 2021

Share "Description of a simple experiment of a TDMA synchronisation principle"

Copied!
24
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

principle

Citation for published version (APA):

Sales, M. A. O. (1977). Description of a simple experiment of a TDMA synchronisation principle. Technische Hogeschool Eindhoven.

Document status and date: Published: 01/01/1977

Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne Take down policy

If you believe that this document breaches copyright please contact us at: openaccess@tue.nl

providing details and we will investigate your claim.

(2)

Vakgroep Communicatiesystemen

Description of a simple experiment of a TDMA synchronisation principle.

M.A.O. Sales

coach: J.A. Greefkes

(3)

Contents O. Summary 1. Introduction

2. Explanation of the proposed system

2.1. Features of the system presented ~n (1) 2.2. The principle of the system

2.3. Description of the synchronisation procedure 2.4. Description of the setting up of communication 3. Circuit of a simplified ground station

3.1. Introduction

3.2. Block diagram of the circuit

3.2.1. Time generation/reception (Fig. 4) 3.2.2. Delay control/transmitter (Fig. 6) 4. Experiment 5. Conclusion 6. Possible extension 7. References Appendix I Figures 2 3 4 4 4 5 6 6 6 6 7 7 8 9 ',9 10 11

(4)

O. Sunnnary

A principle for aquisition of synchronisation in a TDMA system via satellite relay is described and a simplified connnunication between two stations is performed. The circuit of the simplified station is presented. Some possible extensions of the simplified station are suggested.

(5)

I. Introduction

Fig. 1 shows a system in which a number of ground stations (up to a limit of nearly 240) can communicate via satellite relay, all the control of the system being done by a Master Station. In the figure, only two ground stations are shown.

Each station 1S at a different distance from the satellite, implying different propagation times To' T1, T

2.

It is well known that those distances are functions of time, making the problem of synchronism in such a system more complicated.

The purpose of this work is to describe an idea presented 1n (I) for solving the above problem and to perform a very simple experiment applying that principle.

The circuits were implemented in standard printed circuits and using W1re wrapping.

Data sheets about the integrated circuits were obtained from the Signetics Integrated Circuits 1976.

(I) Satellite Communication System for Multidestination, Greefkes, J.A., Dijk, J. and Maanders, E.J.

(6)

2. Explanation of the proposed system

In this section we explain the ideas presented in (1).

The ma~n objective of this system is to establish a data communication network with the following features:

- maximum total bit rate 2.048 Mb/sec.

simple and -consequently- low cost ground stations.

- a processor in one of the stations (Master Station), controls the system by transmitting the correct transmission moment of each ground station and -additionally- operating as a switching center.

- the receivers at the ground stations will be synchronised by the Master Station. - it is possible to have about 240 ground stations taking part in the system,

although only 30 communications can be set simultaneously.

The time format used in the system is shown in Fig. 2.

To each station is assigned one· frame (FR) consisting of 32 time slots (TS) , each time slot having 8 bits. Certainly, besides transmitting data the system should be able to transmit voi~e over the time slots. This requires then that the frame duration be 125 ~sec and -as a consequence- the superframe duration 256 x 125 ~sec = 32 msec.

All ground stations will be synchronised by a code word of 8 bits transmitted

by

the Master Station during TS O. In each frame, TS 16 ~s left for exchange of signalling between the Master Station and each of the ground stations. All ground stations are at different distances from the Master Station and ---these distances are varying with time. How is it possible to make sure that

t.he signalling information will arrive at the satellite in TS 16?

The solution proposed is that the transmission moment of each ground station is ~~ntrolled by the Master Station in such a wa~ that signalling information will arrive at the satellite in TS 16. When the distance between the satellite

and

the ground station changes, the transmission moment is also changed. One fundamental information pointed out in (1) is that these changes in the trans-mission moment do not need to be continuous, since the speed of distance

(7)

variation ~s very low. In this way it is possible to keep all the stations with the correct moment of transmission.

Let us assume that all the stations are switched on, but they are not synchronised. Then:

a. The master station transmits a certain code word (sync. word) during time slot O. This information arrives at the satellite, is regenerated and re-transmitted using another frequency. The master station receives this code word and is able to determine the time delay between the satellite and the master station. With this information, the master station can calculate the delay it should introduce in its receiver such that if information arrives synchronised at the satellite, will also arrive synchronised in the master station. So, from now on we can consider that synchronism with respect to the satellite is

sufficient.

b. Each ground station also receives the synchronisation word and gets synchro-nised with it (that means that clock extraction and reset of time slot

counter ~s made). After this, the ground station selects the signalling channel (TS 16).

c. The master station starts to transmit its address during frame 0 of the superframe (consisting of 256 frames). Each ground station recognizes this address and resets the frame counter.

d. The master station transmits the address of ground station 1 during TS 16 of FR 1 for three successive superframes.

e. When ground station 1 recognizes its own address, it returns this information to the master station, as soon as possible. From this, the master station measures the propagation time delay relative to ground station 1, which ~s

stored in a register. The correct transmission moment for this ground station is calculated in such a way that T8 16 of ground station 1 will arrive at the satellite in TS 16 of the satellite time format. This

information will be transmitted to the ground station during T8 16 of FR 1, using 8 bits (time slot delay given by 5 bits and bit delay given by 3 bits). f. The master station transmits the address of ground station 2 during T8 16

of FR 2 for three successive superframes. The process goes on like described above until all slave stations are synchronised in both directions.

(8)

a. Ground station 1 makes a call by transmitting its o~ address ~n TS 16, FR 0, SF N and the required address in TS 16, FR 0, SF N+1.

b. Master station calls ground station 1 by transmitting the address of ground station 1, followed by information regarding the correct transmission moment and what time slot should be selected. All these informations go during TS 16 of FR 0 of successive superframes.

c. The correct transmission moment for each ground station is under control of the master station and is periodically corrected and recorded, even when no communication occurs.

d. If ground station 1 is out of order for some time, it starts aga~n receiving the correct transmission moment, periodically transmitted by the master station and transmitting its own address during TS 16, FR

°

(for station 2, TS 16 and FR 1; for station 3, TS 16 and FR 2 and so on).

3. Circuit of a simplified ground station

3.1. Introduction

The ground station we built is one which does not perform clock extraction from the incoming signal (then an external clock is provided) and which does not have the receiving signalling logic. The tasK of the signalling logic is to detect and store the delay that should be introduced and the time slot that should be

selected. Both informations corne from the master station as explained in 2.3. and 2.4.

Besides this, the simplified ground station does not have the circuit for providing the initialisation described in 2.3-b and 2.3-c. It is desirable to have this situation in the experiment we wanted to carry out, since this lack of initialisation represents a time delay between two ground stations.

A simplified block diagram is shown in Fig. 2.

We

divided the ground station in two parts: one which makes the time generation / receiving and tbe other which makes the control of transmission delay and the transmission itself. These two parts were built in different standard printed circuits. One reason for this

(9)

separation ~s that we have ~n each part space left for addition of more

integrated circuits when the real ground station (without the simplifications mentioned in 3.1) is to be built.

A more detailed block diagram for the time generator and receiver is shown in Fig. 3. In Fig. 4 we see some related wave forms.

A more detailed block diagram for the transmitter ~s shown ~n Fig. 5. Also in Fig. 6 we see some related wave forms.

Time slots were generated using an 8 bit counter and comparing the five most significant bits (QD'

Ql, ... ,

Q~) with a reference. Then, the duration of the time slot will be 8 clock periods.

In the case of TS

a

and TS 16 the reference with which the input of the counter ~s compared, is fixed (00000 and 01000 respectively).

For TS N, TS N+16, TS Nil and TS N+N" this reference can be changed according to adequate switches. It ~s not difficult to verify that we can obtain TS N+16 from TS N inverting one of the outputs of the time slot counter. To obtain TS N+N" we used a five bit adder before making the comparison.

To obtain FR

a

we used FR 0 as the clock of an 8 bit counter and a 8 bit com-parator. The input signal contains data and signalling information. It is known that signalling will be received ~n TS 16 and data in TS N (which will be

chosen by master station). Then, to obtain signalling (RX) and data to be decoded we just have to gate the input with TS 16 and TS N. We used NAND gates and

inverters for this purpose.

The method adopted for introducing the delay control is the following. We devided the delay in two parts: time slot delay given by 5 bits (switch Nil) and bit delay given by three bits of switch N.

Time slot delay is obtained in the following way: information is stored in a SIPO (series input, parallel output) register during TS (N+16). For this' purpose we generated the clock pulse CP, from CK and TS (N+16).

After CP, information will be at the parallel output of the SIPO r~gister, as shown in Fig. 6. This information is loaded into a PISO (parallel input, series output) register and is sent forward during TS (N+N"). This series

(10)

output is given by

Q

H in Fig. 6.

The control of the second register is very much simplified by entering with a

a

(zero) in the input correspondent to the first bit and using the serial in-put to enter with the last bit.

Thus, during time slot T8 N+N" the parallel input information ~s loaded. At the input appears a "0". When the first clock arrives we have the first bit at the output; when the second clock rises, we have the 2nd bit and so on. In this way, the output

Q

H is obtained.

In a very similar way Q'H is obtained, the only difference being that this information is transmitted only during FR 0 and T8 16. The theoretical time slot delay is "[

=

[(N+N") - (N+16)]l'I

=

(N"-16)l'I where l'I

=

time slot duration. Bit delay is introduced by entering a 8IPO register and connecting the out-puts of this register

1 ...

07'

Fig. 6) to the inputs of an 8 input/multi-plexer.

By selecting the input (by means of 3 bits), the output ~ can be delayed from

o

up to 7 bit periods.

The transmitter must receive data information during T8 N+16 of any frame and signalling information only during T8 16 of FR O.

If. Experiment

As mentioned before, the experiment we carried out was a very simple one, but ~s sufficient to show that the principle of introducing a controlled delay in the transmission ~s sufficient for obtaining synchronism between stations. In Figs. 8 and 9 is shown the layout for one station.

The set up of the experiment is shown in Fig. 10.

As already mentioned, these simplified ground stations are not reset by the master station. Then, every time we switch the equipment on, the counters in

the time slot and frame counters have a different initial condition and, therefore, time slots are not synchronised.

In fact this is equivalent to having the time slot and frame counter reset by the master station and an unknown delay between the two stations.

It was shown that by choosing properly the input ~n switch N" it is possible to synchronise the two stations and therefore to transmit data from one to the other.

The input to the switch N", which is g~ven to control the delay, was obtained by examining the incoming signal and T8 N with the oscilloscope.

(11)

When the delay is such that these two time slots coincide~ then synchronism was obtained between the two stations.

5. Conclusion

We verified that controlling the transmission moment of station 1 (and using one determined time slot) we always find the situation where the data word transmitted by station 1 is received in station 2.

As mentioned in 3.2~ there is one only case where we have problem: when TS N+N" :: TS N+16~ which corresponds to a delay zero or integer multiple of

the frame period. The

~~e~

applied to introduce the delay presents problems when we have this situation.

This problem can be solved in two different ways:

a. when this situation is approached~ a known delay ~s introduced automatically or the information is stored in a different time slot.

b. change the delay (switch Nil) making the information arrive at the satellite in another time slot.

Certainly this second solution implies in more complexity of the soft ware of the master station~ but avoids any additional hardware in the ground station. Since a ground station as simple as possible is desired, we prefer the second solution. It must be noticed that those problems can also happen when TS N+N" and TS N+16 are neighbours.

6. Possible extension

To finish the ground station completely there are still some circuits to be done. Task of these circuits is to make the signalling decoding as described in 2.3 and 2.4.

One suggestion of the circuit of signalling logic is shown in a block diagram form in Fig. 11. One very simple circuit consisting of only one integrated circuit which will transmit the address of the ground station (see 2.3.e) is presented in Fig. 12.

In the circuit of Fig. 11, when the synchronism word is received~ a pulse (a) will clear the time slot counter of Fig. 4; when the address of the master

station is received~ a pulse (S) with one clock period will be the input of a Schmidt-Trigger; the latter will produce a pulse with duration of one frame~

(12)

ground station ~s re~eived during TS 16 of its frame number, a pulse ~s sent to the transmitter of the address of ground station (Fig. 12).

In the circuit of Fig. 12, when a pulse with one clock period duration appears at pin 15, then the address of the ground station is loaded and sent serially by pin 13.

7. References

For the purpose of understanding the used principle and being introduced to additional literature, the essential paper is:

1. Satellite Conwunication System for Multidestinationll ,

(13)

Appendix I. Time slot delay

The problem can be put in this way: given the information ~n TS (N+I6) TS(N+I6)

We have to store it and transmit aga~n ~n TS (N+N") (delay)

TS (N+N") Essentially the same can be said about signalling.

We see from the above that the delay will be given by (N+N"-N-I6)=(N"-16)=N' time slot duration. We expect problems when TS (N+I6) and TS (N+N"} occur at the same time or are subsequent time slots. This can be understood realising that information can not be stored in a register (SIPO) and loaded to another register (PISO) at the same time. When TS (N+I6) and TS (N+N") are subsequent than race conditions can occur.

These problems were confirmed in practice and two suggestions for avoiding it are given in section 5. A diagram of the time slot delay circuit is shown below:

c.P

SIPo (7'110&\) "T J.o

Q

Ii

where:

cp-6

clock pulses occurring during TS (N+I6) Io-input information coming during TS (N+I6) A, , H-parallel outputs of the SIPO register A", , H"-parallel inputs of the PISO register S ser~es input of PISO register

L shift/load control of the PISO register

Q

H delayed output.

Using the above idea of entering the first parallel bit position (H') with "0" and the last bit entering the series input, the control of those registers ~s very simplified.

Of course the circuit for delaying signalling information is the same, except for the control time slots.

(14)

(iROUND STATION#:"j

s-rATION

C;~OUND

(15)

SF: SOPER,FA. AM£

FR:

FP..f.\I'II\£ TS: TfNt:: S Lo

T

\ \ \ \ \ \ \ \ #30

#31,

ffl<

I t \ I I I I I I

,

I I , I I I

I

/ / \ / / / / / / / / I

( I

, I

/ / / /

#0

#i

#2..

#ls

+,-(--tl-':""---+-I_-J.'~/L ) 11/

<

'> TIMe SLOT I / I I OITS~

(16)

1/

[

TSLOT --3l- TS 0 '" s~LE.CTOR TIME -?Ts16 ~T.5 rJ ~j;;N~A.ATID/'oJ i-? TS tJ-rl6

I

TS!.oT

I

,. .. -'7 15 Nil tJtNf[ i:)ELAY

1-7

Ts

>

REC.e:,,,

E-R. AND TI HE 4S /'leA-ATlON }

I

j

I - - - ? ou

TPU

T

i

BIT DELAY CONTRoL DELA '(

Ck

TSL.oT IS Nil 15 NtNI

&

'-5 ",h.. 16 /-Jeui)EK (Ix

..

~

..

---r--~__,~

c." )

TI0Pu ,- - 0 1----'-';>0.;.... 5IGtJALLl rJ

4

1<-.)( Tsl.6

L

-PI<; .. 3 _ S rM~LiFI E"cJ) BLaC k

---_._.

__

.

(17)

Ts 0

k<-

SS,bf'

-?oj

.

._..

. _

---·--~-fj_---J,..---i s N - SAME AS ABO'Jf: BUi W1ff-j R.t=:LATtv!:: FOSt T,ON C.~jAN4E.ABL£

TS N+IG_ II 1I H II 1\ 1I

wIlli A Fl xED PoSITION R,.CLA TIVE:.

II II (AN'D TO IS N 01= 5'8.'3 i"SEc:.) T~ Nil ..::> _ -SAME _PI'.:.

c.,

AS ABoVE I BuT 1 0 Ts N.

ANY POSt TIO,..J

FiZo

1<

32. '\'T)SE.C ---~

_ _11

5

5r---JIl'---.

t:"r~ 5_ WAVEFORMS R.ELAiItD T-o THE C.'IZC.VIT

(18)

71/

5

t

1'0" II \ ~. 2.. 3 4 5 /0 'J /2.. '4 PISD xc(j 6

r-L-/

All

8>

l

t~;t

-rs

Ffl.. Cf<. /6 ()

~

1=1\0 --",!>PI Nlb-~ ..51C}r1(Tx.

,

i

J----'......-

~

~

""i!<:=:---- ...:-_ _TSI.OT /)EI.A

r

---'=:;-'O-~ ...

(19)

, N1"lb

-1

L

p .N£' E

SLllJLfUu:-uLn

_

1 1 1 1

-'l_o_I-I~I~I-I_o::._..+_\

._ _

---'~l---I[,---o

_ _

L

_ _

I~~L-J

,/

_ _ _ _

ILrI--'l~o

__

~.

_ _ _ _ _ _

---In-f[~o

_

_ _ _---11.-1

- - - -

--JI~---f

- - - - -

J

a

4" F E :1> C B A S

,

(20)

(.ofl1P. COl-1P

e.cl,yp

GJ!1F.

/311 r.sLor FIMME

])[;l.Aj coIJNTi£Il. CJ)()fIlTER..

,SloT

bELAy

7--75

IG

(f\J

II)

74/63

'1321

74/63

74&5

142.B3

COMP. C.oMP:

nLoT

FRAME: FAfJ~ A])j)etZ

a~JNrEfl.. cou;t/TfA

INP<~

(21)

1

CJ<.

sr:PO 74166 PI50

3

Fe. 0

IS /6

I /1.

7411

7400

-SIPO 74/66 PISO

'14rSl

'fF,'---t5

e!J...

DIJTPu

r

M!JX FtZoM 131, DELAY

(22)

.

I

_~l

OSC.ILL05-~

-LOPi;"

I

I

S1ATIO~ STATION

I

~i

----

#2-

I

I

WORb .

I

o~TPOT INPUT ....

I

G ENt:RAToR.

I

E.,oJC.(TX)..-

...

..- I

/T· J7/r;

_ . _

-F'I<41.0 _ SE-l U P 0 F THE:

(23)

-T4 77c r/

z)

!

-: I

i

r '/

,

It .I, • I{\ II

,1'

I

1

II

,

H-

'--,.

r3'>

A=B

H-

4=8' A==8

2..x.1

811

2x4

$/T 2)(4 Bir

24

C.0 MPtll<..A,0

R..

C.oMHI IZIJTi~tz

~"'l'-lPAR.ATaR.. I

~

1 t t t t 1 t t

It/~ttTt1'1\ 1t'tll/~ft

S'ftJC- WO~j) 11Di)f{£55 lOP ADD~E.~S o f

/VIA·STeI<. sTarloN Ci-f!..OUND SiAnON

,

V ~

'1/0<.

CLG.AR,. .5CH,I-11i) T

I

'V

T5LOT c.ovr,jTE~ T/vc;CiCR.

S

T 14

SI"//FT'

lFt 4

L.l ) C.LtAR FIU1N}£ DF /1::';';f\../.~., or- /~j)Dk.

fR.

1

CcVNTEI{

(F/::;4)

C;~c<.: Nj) •.s

i'\-?-&,

(r-/0.

I Q.B~ ~I-J /.S

16

...

2x1

f)IT c.--3' 4-L

-QJ)~ c.c.'I'1I-'f.!Rnf»~+L. G~-? ~L Qs-?'- ~L ~-::'J' ""'f-L QI~

f+'L

D ... L t{ FR.of-1 ~I\AIv1£' COlirJi[l{

r-,

L14 )

13

71&5

(24)

/ 2. ~3

A

j)1JI\.G..S5

4

0;:- Cj!:cx)r"fiJ ~~ ST.4 77ut-! 10 // /2, ~

1'1

P:Z:SO IS ,51-liFT/LoAD (ji(oH 9-ft

7

ck 6 L 74/66 Si0N, Lo4 /

C.)

riC; f /

Referenties

GERELATEERDE DOCUMENTEN

In this report, the experimental work for establishing a wireless communication between the ros nodes running on the ground station computer (master) and the companion com-

In de figuur lijkt het dat de afdalende delen van de trap steiler zijn dan de maximale daling van de overkapping. 4p 11 Laat met een berekening aan de hand van de formule zien

Als een kandidaat gerekend heeft met de bijbehorende negatieve waarden voor de daling, hiervoor geen scorepunten in

De oplossing en zeer veel andere werkbladen om gratis te

We assume that the actual arrival times a actual i have a deviation from the scheduled arrival time according to a normal distribution with average 0 and standard deviation 3

 In aansluiting op het vorige punt wordt er aandacht gevraagd voor het governance- aspect: moet er niet in samenspraak met stakeholders worden vastgesteld wat de vraag is, wat

Eerste auteur, jaar van publicatie Type onderzoek, bewijsklasse, follow-up duur Aantal patiënten Patiëntkenmerken Interventie en vergelijkende behandeling

Ook als woorden goed gelezen worden, blijven sommige leerlingen alle woorden letter voor letter lezen en komen ze niet tot directe woordherkenning.. Het leestempo moet dan