IEEE Transactions on Power Systems,Vol.11,No.2,May 1996 917 DETERMINING THE CHANNEL CAPACITY IN SCADA SYSTEMS USING POLLING PROTOCOLS. Joaquin Lugue,Isabel Gomez,Jose I.Escudero Department of Electronic Technology University of Seville Abstract.This article presents a method for that can share the same link thus will be that amount that calculating the capacity of a multi-point communications gives the lowest cost (i.e.the maximum number of channel when a polling protocol is used.Both an exact solution remotes)without violating the communication time and an approximate but easier-to-use solution are obtained. requirementes.In practice,the calculation of this optimum With the results derived here,the number of remote stations number of remotes is carried out by informal,heuristic that can communicate with the control center over a single channel can be determined.The analytical results are methods,often depending upon the experience of the compared to the traffic results measured experimentally. system's designer. In this paper we will try to demonstrate a formal I.INTRODUCTION method for responding to the following question:how many RTU's can share a link without degrading system The use of SCADA (Supervisory Control And performance?Obviously,the answer will depend upon Data Acquisition)systems is a well-established reality in various parameters such as link velocity,message length, the operation of electrical power networks.Frequently the amount of information generated by each RTU,etc.But these systems have a distributed structure with a star it will also depend significantly upon the communications topology in which some remote stations(RTU:Remote protocol used to communicate between the control center Terminal Units)comunicate with a control center,which and the remotes [3].In many control centers,this protocol sends them the network status and receives the pertinent takes a question-answer form [4][5].The control center commands [1].To reduce the connection costs for the polls the first remote on the link;if the remote has communications required for this type of architecture, information to send,it does so,and if not,it sends a null communications between the control center and each RTU message.The control center goes on the poll the second is not point-to-point,but rather the control center splits remote and so on successively until it has contacted all the several communications lines so that several remotes remotes on the link,at which time it starts over again with communicate over each one.Thus,the connection is multi- the first RTU.While keeping in mind the full extent of this point [2].Obviously,the greater the number of remotes procedure,we will concentrate our study on a protocol that share the same communications link,the fewer total such as the one described. number of connections that will be required and,therefore, the lower the cost of the communications infrastructure that II.GENERATING THE INFORMATION will be required for SCADA functions.On the other hand, however,by increasing the number of remote stations For the purposes of our study,we will consider sharing a link,the time assigned to each one will be less, that an RTU transmits three types of information to the and thus the time required to communicate all the RTU's control center [6],two of which are done cyclically,and information will increase.The optimum number of remotes the third,sporadically.The first type of cyclic information (typically information about analogical and digital network 95 SM 557-9 PWRS A paper recommended and approved measurements)we shall call "measurements",and it by the IEEE Power System Engineering committee.of the contains a perishable image of the electrical network;which IEEE Power Engineering Society for presentation at is to say that once a certain amount of time has passed,if the 1995 IEEE/PES Summer Meeting,July 23-27,1995, it has not been able to transmit,it will become invalid and Portland,OR.Manuscript submitted December 14,1994; made available for printing April 28,1995. will be substituted for other equivalent,but more recent, information.This type of information is by far the most common.In contrast,the second type of cyclic information (for example,commands to the remotes,control information,etc.),which we shall call "general"messages, does not lose its validity over time and is not substituted for more recent information.Finally,we will consider that the 0885-895096$05.00©1995IEEE
918 remote is capable of generating sporadic and spontaneous messages-which we shall call "incidents"-with urgent non-nut response information (usually alarms and incidents that have om occurred in the network)that should be sent on to the control center as quickly as possible.Let us also consider that each RTU generates measurement messages in accordance with an exponential distribution with average Tm,or in other words,each RTU generates,on average, one measurement message every T seconds.Similarly,we will consider that general messages are generated exponentially in each remote with an average of T seconds.Finally we will assume that incident messages are very rare and do not affect normal channel traffic.Keeping all this in mind,we will state that each RTU generates messages exponentially with an average T,,shown as 11+1① Fig.1.Temporal diagram of the polling protocol. 7 T T.T or 1 2P+2T.+2D,) while the null response cycle will have a duration of III.PROTOCOL BEHAVIOR WITH LITTLE TRAFFIC bn-PtM +2T。+2D,(4 。 The polling protocol under consideration alternates cycles of questions whose response will be a message with If in a complete polling cyele to all (N)remotes on the link information,with other cycles whose response is a null N,null responses and thus N-N,responses containing message.In any case,in order to study the protocol's information are received,the total time elapsed will be behavior,the following parameters must be considered: T-Np+(N-N)tpm(5) P: Number of bits in the polling message and the nul message. Now let us see what the protocol's capacity is.We M: Number of bits in the messages with know that each remote generates on average one message information (we will assume that they all with M bits every T,seconds,meaning that the total traffic have the same length). generated on a line with N remotes is N.M/T,bps.If we Cp:. Capacity in bits per second (bps)of the initially assume that all the messages generated by the physical link used. remote are finally transmitted by the protocol,which occurs To: Switching time required to change from when the channel has little traffic,the capacity of the receiving or idle state to transmission protocol will obviously be state. D Propagation delay caused by the physical C*NM⑥ medium. T In Fig.1 these parameters can be seen in a typical and remembering that the capacity of the physical medium example which has two cycles of null responses and one is Cp,we can calculate the protocol's efficiency by which contains information.It is easy to deduce that the time elapsed in a response cycle with information is: s+.CN4仞 Cp CT
919 IV.PROTOCOL BEHAVIOR:GENERAL SOLUTION The previous equation is only valid when the n.-∑k-10p0阳(12) traffic over the channel is very low,and thus all the information generated by the RTU's can be effectively transported by the protocol.Nevertheless,when the traffic increases,the transmission of a particular message will require a longer wait,thereby forcing into action the r. mechanism which renews the perishable information.New measurement information is generated by an RTU before the previous data could be sent;hence this new information substitutes the old whose time limit has expired and will thus never be sent.We call this phenomenon message This summation can be shown to equal collision,and we will assume that it affects all the messages generated in the RTU.(This is a good approximation,since the majority of the messages ne=e -1.④ generated are perishable,and only a few are general T, messages.)Let us now calculate the effect the collision process has on the behavior of the protocol. The number of messages generated by a remote and Each remote generates messages exponentially modified by the collision effect in the queue-i.e.the with an average time of T.This is a typically Poisson number of actual messages generated in interval T,-is process for which we know [7]that the probability that it will generate k messages in time T is given by equation %17a吗 p(k) This is equivalent to saying that each remote generates an actual message every Tseconds.This time can be related to T and T by the equation On average,the number of messages generated by a remote during a polling cycle to all the remotes will be n,-∑kp(() T 1-e Substituting the T,generation interval for the actual (10) generation interval in the efficiency equation(7)we obtain s=N M N M-e云.(7 which can be shown to equal C Tre CoT n,= Substituting T for equation(5)we obtain T NeN-N NM S=- (1-e But not all the messages generated are transmitted since Ce [N tp+(N-N)t] some of them are eliminated by collisions with others in the message queue.The average.number of messages that collide and are therefore eliminated in a time period T is Everything is known in this expression except N.But if in given by the following equation a cycle with duration T,(N-N)informational messages
920 are transmitted,the capacity of the protocol will be equation for the efficiency shown in (21)will coincide with C-(N-N)M the general solution both for low and heavy traffic (19) situations,with some expected deviation in the approximate T, solution for medium traffic loads across the channel.The advantage of using the approximate solution lies in the possibility of obtaining results analytically. and the efficiency is The protocol studied here was implemented,and different measurements of interest were taken of it.Figure (N-N)M S (20) 2 shows the dependence of the efficiency in relation to the Cp Cp [N Ipp+(N-N )ta] number of remotes,for a set of typical parameter values (speed:600 bps;switching time:10 milliseconds; measurement message:380 bits;polling message:60 bits; propagation delay:0.3 milliseconds;generation interval for The exact analytical solution for the protocol described is measurement messages:4 seconds).In this diagram the obtained from the solution of the system of equations points obtained experimentally and the exact and formed by equations (18)and (20)which has two approximate solutions can be seen,which confirm the unknowns,S and N..Unfortunately,this is a complex results obtained above. formula,which must be solved in each case by numerical methods VI.MEASUREMENT UPDATE CYCLE V.PROTOCOL BEHAVIOR:APPROXIMATE While efficiency is a valuable measurement for SOLUTION evaluating the capabilities of a protocol,from the user's point of view other measurements are more desirable.The Nevertheless,an approximation exists which is first one of these tells us what the cycle is for updating the valid for a large number of applications.This mesaurement messages.Though nominally each remote approximation consists of assuming that a polling cycle to sends a measurement message every T seconds,due to the all the remotes always obtains a response containing effect of the collisions,the actual cycleTme could be information from all of them,i.e.that assuming that N.=0. different, especially in situations where the link is Under these conditions (18)becomes saturated.To calculate it,let us consider a time period T. The total number of messages that arrive at the control center within this interval from the N remotes,will be s-M(1-e (21) N,=N.T/T,and the number of measurement messages N=N.T/T Since the general messages are not affected by the collision process,the actual generation of this type of message matches the nominal amount,by.which we can This approximate equation of efficiency (S')will coincide with the exact solution (S)when the channel is saturated, ·Simulated ncy or in other words,when it has a heavy traffic load.But, Exacl both expressions also coincide when the number of remotes is low,and thus the traffic is light.In effect,one only has 0.8 to remember that for small x values,the following approximation can be made: 0.6 1-erax.(22 Applying this to (18)and (21),wo verify that for small numbers of remotes,we can state the following: 13写 7911131517.192123252729 S-sl=NM (23) Numbor of Romotos 11 CT, Fig.2.Protocol efficieney. which on the other hand matches the value of S in(7)for unsaturated channels.In summary,the approximate
921 ·Simulated hand,by the time used to transmit the incident from the Moasure 一pproximate remote to the control center.Since the incident can occur --Exact at any moment,we will assume that with a uniform distribution,the average wait time will be T=T /2,where T was the duration of the polling question-and-answer cycle to all of the remotes.On the other hand,the time required for the transmission of an incident message from a remote to the control center will be given by 0,总m 013679113151719212公25279 Number of Remotes Therefore,the delay in the transmission of an incident is Fig.3.Measurement refresh cycle. given by +T+Dp D M.(28) state that the number of messages received by the control 2 C: center in T seconds will be N.=N.T/T.Obviously, N=N+N or expressed another way,N=N-N Based on the above,the measurement refresh cycle is given by (24) In the previous equation the only unknown value is T which can be obtained by combining equations (5)and NTNT NT TN.N-N NI.NT 1-1 GA) 1 (20).The resulting expression is(29), N M tpp T,M-sc,m-n） ,(29) Remembering (17)we obtain(25) 器 25 substituting this result,we find(30) N M tpp which,when substituted into (24)gives (26) 0C4万0总m 1 T." (26 Here again,either the general or the approximate solution NM T can be used for the value of efficiency(S),according to the desired objectives.Figure 4 shows both the theoretical and experimental values obtained for the incident transmission In (26)the value for efficiency (S)can be obtained from delay. either the general or the approximate solution,according to whether greater precision or ease of calculation is VIII.CONCLUSIONS preferred.Figure 3 shows the values for the measurement update cycle according to the theoretical results derived In this paper,equations have been obtained which here,and are compared to the experimentally obtained relate the efficiency,the measurement update cycle and the results. incident transmission to the number of remotes on a link operated with a polling protocol.Using these results,the VII.INCIDENT TRANSMISSION DELAY maximum number of remotes that can be placed on a multi- point link connected to a control center can be determined. Let us now look at the delay in transmitting an In effect,this maximum value for the number of remotes incident.This delay is composed of,on the one hand,the will be that amount at which the efficiency,the average RTU wait time from the moment the incident measurement update cycles and the incident transmission occurs until transmission of it begins,and on the other delays stay within the bounds of the permissible values
922 ·Simulated X.REFERENCES -Approdmate [1] Dennis J.Gaushell and Henry T.Darlington, "Supervisory Control and Data Acquisition" Proceedings of the IEEE,vol.75,no.12, December1987,pp.1645-1658. [2] D.J. Gaushell and J.H.Noland, "Telecommunication Options",IEEE Tutorial Course,Fundamentals of Supervisory Systems, 1991,p.35. 357911131517192123.252729 [3] Isabel Gomez,Joaquin Luque and Jose I. Numbor of Remotes Escudero,"Medium Access Control Protocols for Fig.4.Incident transmission delay. Electrical Power Network Control",Proceedings of the 1992 Bilkent International Conference on Lightwave Technology and Communications, Bilkent University,Ankara,Turkey,July 1992. necessary for proper network operation. pp.23-29. IX.BIOGRAPHIES [4④ L.Krishnan and W.Zimmer(Editors),"Polling Joaquin Luque received his degree in Industrial Engineering in 1980 and vs.Event Reporting.What are the Tradeoffs?", his Doctorate in Industrial Engineering in 1986 from the University of Integrated Network Management Il,Elsevier Seville (Spain).Since 1980,he has worked for several companies in the Science Publishers B.V.(North-Holland),1991, area of SCADA systems for electrical networks,participating in some of p.339 the primary EMS projects in Spain.He is currently a professor of eleetronic engineering at the University in Seville,and he isa member of the IEEE. [S Prasad Raja,Guevara Noubir,Luis Ruiz,Jean Hernandez and Jean-Dominique Decotigne, Isabel Gomez received her Physies(Electronics)degree in 1989 from the "Analysis of Polling Protocols for Fieldbus University of Seville (Spain).She has beena professor of eleetronie Networks",ACM SIGCOMM,Computer engincering at the University in Seville since 1990,where she is doing Communication Review,pp.69-90. research on problems in computer communications for the control of electrical power networks. I6) H.Lee Smith and Wayne R.Block,"RTUs Slave Jose I.Escudero received his Physics degree in1979 from the University for Supervisory Systems",IEEE Computer of Seville(Spain).Since then,he has held several teaching positions.He Applications in Power,January 1993. has been a professor of electronic engineering at the University in Seville since 1989.Mr.Escudero has focused his research activity on the study of computer nctwork performance. [) Mischa Schwartz,Telecommunications Networks Protocols,Modeling and Analysis,Addison- Wesley,1988,p.24