M. J. Johnson, "Proof that Timing Requirements of the FDDI Token Ring Protocol are Satisfied", IEEE Trans. Communications, COM-35, No. 6, pp. 620-625, 1987. 17  Home/Search   Document Not in Database   Summary   Related Articles   Check

....TRT TTRT, the difference can be used for sending asynchronous frames. Otherwise, the token is late and no asynchronous frame can be transmitted. Using this scheme of bandwidth allocation, the mean token rotation time is at most TTRT, and the maximum rotation time can never exceed twice the TTRT [17]. 3.1 The simulation model An FDDI network can be considered as a polling system. A polling system is defined as a multiple queue singleserver system. We refer to the time needed to switch service from one queue to the next one as the walk time, and the sum of all these times as ring latency. ....

M.J. Johnson. Proof that timing requirements of the FDDI token ring protocol are satisfied. IEEE Transactions on Communications, 35:620-- 625, June 1987.

....basic idea) of the timed token protocol, adaptable to either a physical or a logical ring, was described. Ulm [3] then studied the protocol proposed by Grow and its performance characteristics. The timing properties of the timed token protocol were first formally analyzed by Johnson and Sevcik in [4, 5] where it is shown that the average token rotation time is bounded by the Target Token Rotation Time (TTRT ) and the maximum token rotation time cannot exceed twice the TTRT . Chen et al. [1, 6, 7, 8] made a detailed study on the timing behavior of the timed token protocol and generalized the upper ....

....nodes, i.e. P n i=1 i . 1 The downstream neighbor of node i is node i 1 if i n else node 1 if i = n; similarly, the upstream neighbor of node i is node i Gamma 1 if i 1 else node n if i = 1. 2 Various overheads possibly involved have been identified by Johnson and Sevcik in [4, 5]. For example, protocol dependent overheads include token capture delay, token transmission delay, etc. 3 3 Timed Token MAC Protocol The basic ideas of the timed token protocol were first presented by Grow [2] With this protocol [20] messages are distinguished into two types: synchronous and ....

[Article contains additional citation context not shown here]

M. Johnson, "Proof that Timing Requirements of the FDDI Token Ring Protocol are Satisfied," IEEE Trans. on Communications, vol. COM-35, pp. 620--625, June 1987.

....in part by the CEC, through Esprit Project 1226 DELTA 4 and by JNICT through Programa Ciencia. which are unperfectly fulfilled by these LANs, if used without special measures. A number of authors have studied problems such as priority inversion [1] probability of meeting estimated access times [2,3], extensions for medium failure resiliency through redundancy [4] potential lack of determinism [5] In reliable real time systems, the fundamental requirement of communications is that there be a bounded and known message delivery latency, in the presence of disturbing factors such as overload ....

Marjory J. Johnson. Proof that timing requirements of the FDDI token ring protocol are satisfied. IEEE Transactions on Communications, 35(6), June 1987.

....requirements of a number of applications, specially in the fault tolerance and real time area, which are unperfectly fulfilled by these LANs, if used without special measures. A number of authors have studied problems such as priority inversion [1] probability of meeting estimated access times [2,3], extensions for medium failure resiliency through redundancy [4] potential lack of determinism [5] In reliable real time systems, the fundamental requirement of communications is that there be a bounded and known message delivery latency, in the presence of disturbing factors such as overload ....

....both classes is prevented after TRT expiration. This methodology of accessing network medium presents timeliness characteristics guaranting that, under normal operation, a token always return to any given station in a time upper bounded by 2 Theta T Opr. This claim has been formally proofed in [3] and furnishes the foundations for a method aiming the detection of incorrect ring activity: Ring recovery is initiated upon TRT has expired twice without any token has been received, i.e. recovery is started when TRT expires and Late Ct 0. Establishment of an expression for inaccessibility ....

Marjory J. Johnson. Proof that timing requirements of the FDDI token ring protocol are satisfied. IEEE Transactions on Communications, 35(6), June 1987.

....This kind of service is also required from the link layer by some higher layer protocols, like TENET[1] that supports real time tra#c generated by the transport layer. The first protocol to support such service requirements was the MAC protocol of the FDDI. FDDI uses the timed token protocol [6], which distinguishes between synchronous and asynchronous packets. Synchronous packets, such as packetized voice or video, are generated at regular intervals and have delivery time constraints. Asynchronous packets, in contrast, are non periodic and have no time constraints. The time token ....

....and have delivery time constraints. Asynchronous packets, in contrast, are non periodic and have no time constraints. The time token protocol guarantees to each source of a synchronous tra#c an average bandwidth and a bounded delivery time. The latter depends on some parameters, mainly the TTRT [6], selected during ring initialization. Supporting synchronous tra#c on the downstream channel is relatively simple. This is because this channel is governed by a single entity, the HE, that can schedule packets for downstream transmission according to any pre determined set of rules. In contrast, ....

M. Johnson. Proof that timing requirements of the FDDI token ring protocol are satisfied. IEEE Transactions on Communications, 35(6):620--625, June 1987.

....neighbor. The sum of all such latencies is known as the ring latency, Theta, i.e. Theta = P i i . Let Delta denote the time needed to transmit a maximumsize asynchronous message. Then = Theta Delta represents the portion of time unavailable for data transmission. It has been shown in [18] that if P i h i TTRT Gamma , each station is guaranteed to see the token at least once in every 2TTRT time. This result is generalized in [9] to the following. Theorem 1 If P i h i TTRT Gamma , the time elapsed between any n consecutive token visits to station i is bounded by n TTRT ....

....allocated to the stations on the way from station j to i, including both j and i. We need the following lemmas to prove our theorem. Lemma 2 ( 9] If the token arrival at station p at time t p (n) is not late, then for any 0 m n, we have t p (n) Gamma t p (m) n Gamma m)TTRT. Lemma 3 ([18]) After initialization, TRT values of all operational stations will monotonically increase in the down link direction, up to and including the station that last received the token. Theorem 6 For any n m 0, the time elapsed between t p (m) and t q (n) is bounded by (n Gamma m 1)TTRT B max ....

M. J. Johnson, "Proof that timing requirements of the FDDI token ring protocol are satisfied," IEEE Transactions on Communications, vol. Com-35, no. 6, pp. 620--625, June 1987.

....of multimedia communication systems. In this paper, issues in meeting time constraints of multimedia traffic in practical FDDI based multimedia communication systems are addressed. There has been extensive research on using FDDI networks for time critical applications. Johnson and Sevcik [10, 19] proved that when the network operates normally, the time elapsed between two consecutive visits of the token at a node is bounded by 2 Delta TTRT . This result was generalized by Algrawal et al. 1] that the time elapsed between any v consecutive visits is bounded by v Delta TTRT , and that the ....

M. J. Johnson. Proof that timing requirements of the FDDI token ring protocols are satisfied. IEEE Transactions on Communications, COM-35(6):620--625, June 1987.

....in FDDI networks are observed only under conditions of high network load. We assume that all voice data is transmitted as synchronous traffic. With the recommended default value for the Target Token Rotation Time (TTRT) of 8 ms [27] an upper bound of the access delay is given by 2 Delta TTRT [31]. Assuming that T access is exponentially distributed, we obtain a truncated Cox distribution Phi(8; 1; 16) for the network access delay. 33 Parameter Description Min. Max. Avg. T net One way network delay 2 36 12 T sndr Processing delay at sender 1 10 2 T rcvr Processing delay at receiver 1 ....

M.J. Johnson. Proof that Timing Requirements of the FDDI Token Ring Protocol Are Satisfied. IEEE Transactions on Communications, COM-35:620--625, June 1987. 99

....The sum of all such latencies is know as ring latency, Theta, i.e. Theta = P i i . Let Delta denote the time needed to transmit a maximum size asynchronous message. Then = Theta Delta represents the potion of time unavailable for synchronous data transmission. It has been shown in [11] that if P i h i TTRT Gamma , each station is guaranteed to see the token at least once in every 2TTRT time. This result is generalized in [6] to the following theorem: Theorem 1 If P i h i TTRT Gamma , the time elapsed between any n consecutive token visits to station i is bounded by ....

M. J. Johnson, "Proof that timing requirements of the FDDI token ring protocol are satisfied," IEEE Transactions on Communications, vol. Com-35, no. 6, pp. 620--625, June 1987.

....token to make one rotation through all the stations in the network. The Synchronous Bandwidth (H i ) of each station: the time guaranteed for transmitting synchronous message every time the station receives the token. The timing properties of the FDDI protocol have been extensively studied. In [10], Johnson and Sevcik formally proved that when the network operates normally, the upper bound on the token rotation time is twice the expected token rotation time, i.e. 2 Delta TTRT . The difference between the maximum token rotation time and the achievable average token rotation time halves an ....

....In this section, we will attempt to further classify SBA schemes according to the strategy they use in allocating the bandwidth, and also demonstrate the form in which these SBA schemes must take in order to maximize the WCAU. We first note that any SBA scheme must satisfy two ba sic constraints [10, 2]: 1. Protocol constraint: the sum of the synchronous capacities allocated to all nodes in the FDDI ring should not be greater than the available portion of TTRT i.e. n P i=1 H i TTRT Gamma . 2. Deadline constraint: the allocation of the synchronous capacities to the nodes must guarantee ....

M. Johnson. Proof that timing requirements of the fddi token ring protocol are satisfied. IEEE Trans. Comput., 35:620-- 625, 1987.

....timer expires. In this way, the time it takes for the token to circle along all the stations can be bounded, thus providing facilities for guaranteed synchronous bandwidth allocation. A number of basic timing properties of FDDI access mechanism have been studied by Jain [25] Johnson [26] and Sevcik and Johnson [36] The global time based mechanism is very difficult to incorporate in analytical polling models. Apart from approximate approaches, e.g. those by Tangemann [42] we are not aware of any analytical results. Still though, the behaviour of such systems can easily be ....

.... the station (transition start timer) just before the server switches to the next downstream station (transition switch) The length of the timer is the so called target token rotation time (ttrt) If all stations obey this timer, the (network) token will cycle around within a precomputable time [26, 36]. When the server arrives in place token and the timer has not yet expired, two things can happen. If the queue of jobs is empty (place buffer) the immediate transition empty fires so that the server starts going to the next station. This transition also takes care of removing all tokens from the ....

M.J. Johnson, "Proof that the Timing Requirements of the FDDI Token Ring Protocol are Satisfied", IEEE Transactions on Communications 35(6), pp.620-- 625, 1987.

....requirements of a number of applications, specially in the fault tolerance and real time area, which are unperfectly fulfilled by these LANs, if used without special measures. A number of authors have studied problems such as priority inversion [1] probability of meeting estimated access times [2,3], extensions for medium failure resiliency through redundancy [4] potential lack of determinism [5] In reliable real time systems, the fundamental requirement of communications is that there be a bounded and known message delivery latency, in the presence of disturbing factors such as overload ....

Marjory J. Johnson. Proof that timing requirements of the FDDI token ring protocol are satisfied. IEEE Transactions on Communications, 35(6), June 1987.

....guaranteed bandwidth so that real time applications are allowed; whereas the asynchronous traffic is controlled by the station timers. Such a timed token scheme was first suggested by Grow [12] Ulm [26] analyzed the influence of different parameters on the utilization of the ring. Johnson [14] and Sevcik and Johnson [22] proved cycle time properties in the FDDI token ring protocol. Although the ability to deal with synchronous traffic is one of the particularities of the FDDI protocol, most of the current implementations of the protocol uses the asynchronous transmission mechanism ....

....can be transmitted when the inter visit time 2 (IT) is less than TTRT. In the remainder of the paper, the original FDDI MAC scheme will be referred to as FDDI with respect to (w.r.t. CT, denoted FDDI CT, and the new one as FDDI w.r.t. IT, denoted FDDI IT. 3 Cycle Time Constraint It is shown in [14, 22] that in FDDI CT, the maximum cycle time is bounded above by 2 times TTRT. We will show in this section that the same property holds for the modified FDDI protocol, viz. FDDI IT. Theorem 3.1 Consider FDDI IT. Unless initiation of ring recovery interrupts the cycle, the maximum cycle time is ....

[Article contains additional citation context not shown here]

M. J. Johnson, "Proof that Timing Requirements of the FDDI Token Ring Protocol are Satisfied", IEEE Trans. Communications, COM-35, No. 6, pp. 620-625, 1987.

....ring protocol to provide priority arbitration in a bus network would be very inefficient, because of the overhead due to passing the token. 4. Some other synonymous terms that researchers use are bandwidth allocation (Ulm 1982) synchronous allocation (Jain 1991) synchronous bandwidth assignments (Johnson 1987), high priority token holding time (Pang and Tobagi 1989) and synchronous capacity (Agrawal et al. 1992, 1993) We use the term synchronous bandwidth in order to be consistent with the current FDDI standard. 5. TTRT can be no more than half of the minimum message deadline because the maximum time ....

....1993) We use the term synchronous bandwidth in order to be consistent with the current FDDI standard. 5. TTRT can be no more than half of the minimum message deadline because the maximum time between consecutive token visits to a node can approach 2TTRT (Agrawal et al. 1992; Chen et al. 1992; Johnson 1987). 6. The synchronous bandwidth allocation schemes listed below assume that there is only one stream of synchronous messages originating at each node. This assumption involves no loss of generality, because a network in which several streams may originate at each node can be transformed into a ....

Johnson, M. J. 1987. Proof that timing requirements of the FDDI token ring protocols are satisfied.

....than expected. In order to guarantee that the deadlines of synchronous messages are met, network parameters such 1 Some other synonymous terms that researchers use are synchronous capacity [1, 2, 4] bandwidth allocation [18] synchronous allocation [7] synchronous bandwidth assignments [10], and high priority token holding time [13] This paper uses the term synchronous bandwidth, in accordance with the most recent version of the FDDI standard. as the synchronous bandwidth, the target token rotation time, and the buffer size must be chosen carefully. ffl The synchronous bandwidth ....

....n and j = 1; 2; s i;j t i;j D i (7) where t i;j is the arrival time and D i is the (relative) deadline. Note that in the above inequality, t i;j and D i are given by the application, but s i;j depends on the synchronous bandwidth allocation and the choice of TTRT. Johnson and Sevcik [10, 15] showed that for the timed token protocol, the maximum amount of time that may pass between two consecutive token arrivals at a node can approach 2TTRT . To satisfy the deadline constraint, it is necessary for a node to have at least one opportunity to send each synchronous message before the ....

[Article contains additional citation context not shown here]

M. J. Johnson. Proof that timing requirements of the FDDI token ring protocols are satisfied. IEEE Transactions on Communications, COM35 (6):620--625, June 1987.

....in FDDI networks are observed only under conditions of high network load. We assume that all voice data is transmitted as synchronous traffic. With the recommended default value for the Target Token Rotation Time (TTRT) of 8 ms [13] an upper bound of the access delay is given by 2 Delta TTRT [16]. Assuming that T Acc is exponentially distributed, we obtain a truncated Cox distribution Phi(8; 1; 16) for the network access delay. The key parameters of a packet voice protocol are the control time at the PVR queue, T V , and the packetization interval, T p . Both T V and T p are constrained ....

M.J. Johnson. Proof that Timing Requirements of the FDDI Token Ring Protocol Are Satisfied. IEEE Transactions on Communications, COM-35:620--625, June 1987.

....the j th message in stream S i , and let s i;j be the time that the transmission of this message is completed. The deadline constraint implies that for i = 1; n and j = 1; 2; s i;j t i;j D i (10) where D i is the relative deadline of messages in stream S i . Johnson and Sevcik [9, 16] have shown that for the timed token protocol, the maximumamount of time that may elapse between two consecutive token arrivals at a node can approach 2TTRT . To satisfy the deadline constraint, it is necessary for a node to have at least one opportunity to send each synchronous message before the ....

M. J. Johnson. Proof that timing requirements of the FDDI token ring protocols are satisfied. IEEE Transactions on Communications, COM-35(6):620--625, June 1987.

....time. The node can transmit its asynchronous messages only if the time interval between the previous token arrival and the current token arrival is less than TTRT. Many researchers studied the access time bounds and other timing properties of the timed token protocol. In particular, Johnson et al. [5, 6] proved that the average token cycle time is bounded by TTRT, and the maximum token cycle time is bounded by 2 Theta TTRT. Agrawal et al. 1, 2] extended Johnson s result and proved that the time elapsed between k consecutive token s visits to a node is bounded by k Theta TTRT. They also ....

....portion T Gamma of TTRT, i.e. P n i=1 H i T Gamma . Violation of the protocol constraint will make the ring unstable and oscillate between claiming and operational. Under the timed token protocol constraint, the following well known result for the MAC protocol is formally proved in [5, 6]. Theorem 1: Johnson and Sevcik s Theorem) For the timed token protocol, the worst case token rotation time the time elapsed between any two consecutive token s visits to a node is bounded by T P n j=1 H j 2 Delta T . 2 A more general result has also been obtained by Agrawal ....

M. J. Johnson, "Proof that timing requirements of the FDDI token ring protocol are satisfied," IEEE Trans. on Commun., vol. COM-35, pp. 620--625, June 1987.

....timing properties of the timed token protocol. 4.1 Timing Properties A key property of the timed token protocol is that it provides each station a bounded access time of 2 Delta TTRT to the transmission medium. This is because the maximum token rotation time has been shown to be 2 Delta TTRT [12,22]. Agrawal et al. generalized this result to provide an upper bound on the time between any v consecutive visits of the token to a given station [2,3] Let t i (l) denote the time when the token makes its l th visit to station i. Theorem 1: From [2,3] For any l 0, v 0, and any station i ....

M. J. Johnson, "Proof that timing requirements of the FDDI token ring protocol are satisfied," IEEE Transactions on Computers, vol. COM-35, no. 6, pp. 620--625, June 1987.

....temporal guarantees for specific channels with a restricted bandwidth. It is sometimes possible to have specific communication channels with some temporal and bandwidth guarantees. That is, for example, the case of ATM (Asynchronous Transfer Mode) networks[6] or local area networks with priorities[9]. Other general data channels can not be considered fully synchronous, they are at best quasi synchronous. With this scenario it is very difficult and too costly (if at all possible) to use a hard real time approach. It would require a resource adequacy policy that would be too expensive to ....

Marjory J. Johnson. Proof that timing requirements of the FDDI token ring protocol are satisfied. IEEE Transactions on Communications, 35(6), June 1987.

....capacity. It can then transmit its asynchronous messages only if the time elapsed since the previous token departure from the same node is less 1 Some other synonymous terms that researchers use are: Bandwidth allocation [66] Synchronous allocation [26] Synchronous bandwidth assignments [29], and High Priority token holding time [46] than the value of TTRT , i.e. only if the token has arrived earlier than expected. Guaranteeing a message deadline implies transmitting the message before its deadline. With a token passing protocol, a node can transmit its message only when it ....

....of the message s deadline. That is, in order to guarantee message deadlines in a token ring network, it is necessary to bound the time between two consecutive visits of the token to a node (called the token rotation time or the access time) The timed token protocol has this property. In [29, 54], Johnson and Sevcik formally proved that when the network operates normally (i.e, in the absence of any failure) the token rotation time between two consecutive visits to a node is bounded by twice the Target Token Rotation Time (i.e. 2 Delta TTRT ) Although the prerequisite of bounded token ....

[Article contains additional citation context not shown here]

M. J. Johnson, "Proof that timing requirements of the FDDI token ring protocols are satisfied," IEEE Trans. Commun. Vol. COM-35.No. 6, pp. 620-625, June 1987.

....at each node. LAN MAC level priorities allow in principle urgent frames to overtake less urgent ones on the network although it has been shown in [2] that the scheme may be upset in real settings. We will not spend too much time with these issues, since they have been studied by a number of authors[1,11,12,13,14]. Handling omission failures The bounded omission degree 7 assumption introduced in An2 is very helpful as the foundation of basic error processing protocols with deterministic termination crucial for real time operation. Other properties of higher level reliable broadcast multicast ....

Marjory J. Johnson. Proof that timing requirements of the FDDI token ring protocol are satisfied. IEEE Transactions on Communications, 35(6), June 1987.

.... the synchronous bandwidth allocation (ST) and the operational target token rotation time (T Opr ) Parameter tuning and the resulting performance of FDDI is discussed extensively in [3] Proofs for the performance guarantees provided and related timing properties of FDDI are presented in [4] and [5]. Unfortunately, the synchronous mode of FDDI is considered an optional feature by manufacturers and it is not implemented in most FDDI equipment. The most obvious manifestation of this lack of support is hardware limitations. In particular, adapters do not possess separate synchronous and ....

M.J. Johnson, "Proof that Timing Requirements of the FDDI Token Ring Protocol are Satisfied," IEEE Transactions on Communications, vol. 35, no. 6, pp. 620-625, June 1987.

No context found.

M. J. Johnson, "Proof that Timing Requirements of the FDDI Token Ring Protocol are Satisfied", IEEE Trans. Communications, COM-35, No. 6, pp. 620-625, 1987. 17

No context found.

M.J. Johnson. Proof that timing requirements of the FDDI token ring protocol are satisfied. IEEE Transactions on Communications, 35(6):620--625, June 1987. 27

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC