N. Shacham and P. McKenny, "Packet Recovery in High-Speed Networks using Coding and Buffer Management ", Proceedings of IEEE INFOCOM '90, San Francisco, CA, June 1990.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....standard for GSM Adaptive Multi Rate (AMR) codec [33] and the North American standard IS 641 [34] ii) Another category of PLC algorithms operate with certain coordination with the transmitter side. Methods based on Forward Error Correction, transmit redundant repair data to recover lost packets [36] [40] Diversity schemes based on Multiple Description Coding (MDC) also fall into the same category [41] 43] Both FEC based and MDC based techniques impose additional delay; however, as opposed to the former, the latter schemes may not necessarily require additional bandwidth. iii) A third ....

Shacham, N.; McKenney, P., "Packet recovery in high-speed networks using coding and bu#er management," Proc. IEEE INFOCOM, vol.1, pp. 124-131, 1990.

....of a phoneme (5 100 ms) Hence, error concealment schemes should be regarded as complementary to, and not substitute for, FEC. FEC techniques can be classified as media independent and media specific. Media independent FEC uses block codes (e.g. based on Reed Solomon [16] or on parity codes [17], 18] to provide redundant information. Each code takes a codeword of # data packets and generate ### additional check packets for the transmission of # packets over the network. Such a code, denoted as an ### ## code, is able to recover all losses in the same block if an only if at least # out ....

N. Shacham and P. M. Kenney, "Packet recovery in high-speed networks using coding and buffer management," in Proc. IEEE Infocom'1990.

....of a phoneme (5 100 ms) Hence, error concealment schemes should be regarded as complementary to, and not substitute for, FEC. FEC techniques can be classified as media independent and media specific. Media independent FEC uses block codes (e.g. based on Reed Solomon [16] or on parity codes [17], 18] to provide redundant information. Each code takes a codeword of k data packets and generate n k additional check packets for the transmission of n packets over the network. Such a code, denoted as an (n, k) code, is able to recover all losses in the same block if an only if at least k out ....

N. Shacham and P. M. Kenney, "Packet recovery in high-speed networks using coding and buffer management," in Proc. IEEE Infocom'1990.

....with redundancies and those without. Adding redundancies at the packet level is not the best for fault tolerance in the Internet because they require considerable increases in bandwidth over non redundant schemes. Typical methods include adding copies of previous frames [18] using parity or FEC [24] codes to protect every n packets by a redundant packet, using FEC to protect only sensitive information in LP coders [3] and piggy backing in a packet a redundant version of some previous packets obtained by a lower bit rate coder [5] In addition, redundant information can be sent to protect ....

N. Shacham and P. McKenney. Packet recovery in high-speed networks using coding and bu#er management. In Proc. of IEEE INFOCOM, pages 124--131, May 1990.

....also random losses on the incoming and outgoing links to the bottleneck node. In particular, we study the tradeoffs mentioned in the previous paragraph. The problem of analyzing loss probabilities due to congestion losses in the presence of redundant packets has been addressed in earlier works [7, 5, 4, 9]. In [7] the authors have used an approximation based on the assumption of independence between consecutive losses, and have shown that redundancy results in a decrease of loss probabilities by 10 to 100. Exact numerical methods based on recursions in [5] led, to an opposite conclusion, i.e. ....

....on the incoming and outgoing links to the bottleneck node. In particular, we study the tradeoffs mentioned in the previous paragraph. The problem of analyzing loss probabilities due to congestion losses in the presence of redundant packets has been addressed in earlier works [7, 5, 4, 9] In [7], the authors have used an approximation based on the assumption of independence between consecutive losses, and have shown that redundancy results in a decrease of loss probabilities by 10 to 100. Exact numerical methods based on recursions in [5] led, to an opposite conclusion, i.e. adding ....

[Article contains additional citation context not shown here]

N. Shacham and P. McKenney, "Packet Recovery in High-Speed Networks Using Coding and Buffer Management", INFOCOM '90, San Francisco, CA, pp. 124-131, 1990.

....length of a phoneme (5 100ms) Hence, error concealment schemes should be regarded as complementary to, and not substitute for, FEC. FEC techniques can be classified as media independent and media specific. Media independent FEC uses block codes (e.g. based on Reed Solomon [15] or on parity codes [16, 17]) to provide redundant information. Ech code takes a codeword of k data pckets and generate n k additional check pckets for the transmission of n pckets over the network. Such a code, denoted as an (n, k) code, is able recover all losses in the same block if an only if at least k out of n packets ....

N. Shacham and P. Mc Kenney, "Packet recovery in high-speed networks using coding and buffer management," in Proc. IEEE Infocom1990.

....redundant information, along with the original information, so that lost data can be recovered, at least partially, from this redundant information. Originally, there was much interest in the provision of media independent FEC using block codes (e.g. based on Reed Solomon [12] or on parity codes [13]) to provide redundant information. Unfortunately, these techniques have the disadvantage of introducing important additional delays (because a source must wait for the entire block of packets before computing and transmitting the redundancy packet) These schemes can therefore be used for ....

N. Shacham and P. Mc, \Packet recovery in high-speed networks using coding and buer management, " in Proc. IEEE Infocom'1990.

....the data link layer to reduce the impact of errors in the wireless connection. In most cases, these codes provide less than perfect protection and some amount of residual errors pass through. The upper level protocol layers employ various block error detection and retransmission schemes (see e.g. [67][39] With FEC redundancy bits are attached to a packet that allow the receiver to correct errors which may occur. In principle, FEC incurs a fixed overhead for every packet, irrespective of the channel conditions. This implies a reduction of the achievable data rate and causes additional ....

Shacham, N., McKenney, P.: "Packet recovery in high-speed networks using coding and buffer management", Proceedings IEEE Infocom'90, San Fransisco, pp. 124-131, May 1990.

....[8] offers the advantages of a significantly smaller average delay and variance, less sensitivity to link utilization variations, and smaller buffer sizes for a given message loss probability due to buffer overflow. A similar idea has been suggested for lost packet recovery in high speed networks [9], 10] These applications also fall into the category of erasure channel coding. 2 Diversity Coding We assume that the reader is familiar with the theory of finite fields. For more information on this subject we refer the reader to one of the standard texts on the subject such as [11] or [12] ....

N. Shacham and P. McKenney, "Packet recovery in high-speed networks using coding and buffer management," Proceeditzgs of IEEE INFOCOM '90, Vol. 1, pp. 124 131, San Francisco, California, June 1990.

....using a number of mechanisms [7] 8] including local repair (interpolation of missing data using the surrounding packets) and interleaving. There has been much interest in the use of packet level FEC for sending redundant information ahead of time to compensate for loss, based on parity codes, [9], 10] Reed Solomon codes [11] and redundant speech codecs [12] 13] 14] 15] Oftentimes, the term FEC is only applied to the traditional channel coding approaches, such as parity and Reed Solomon codes. For purposes of this paper, we define FEC as any mechanism which sends additional infor ....

Nachum Shacham and Paul McKenney, "Packet recovery in high-speed networks using coding and buffer management," in Proceedings of the Conference on Computer Communications (IEEE Infocom), San Francisco, California, June 1990, IEEE, pp. 124--131.

....redundant information, along with the original information, so that lost data can be recovered, at least partially, from this redundant information. Originally, there was much interest in the provision of media independent FEC using block codes (e.g. based on ReedSolomon [13] or on parity codes [14], 15] to provide redundant information. Unfortunately, these techniques have the disadvantage of introducing important additional delays (because a source must wait for the entire block of packets before computing and transmitting the redundancy packet) These schemes can therefore be used for ....

N. Shacham and P. Mc Kenney, "Packet recovery in high-speed networks using coding and buffer management," in Proc. IEEE Infocom'1990.

....communication media, the assumption is made that decoding is mainly impeded by packet loss. Either the packet is present and correct or it is lost. These losses are mainly caused by network congestion and the resultant buffer overflow and queuing delay. In this case, packet level FEC schemes [24] [27] provide an efficient way to fight against losses, although perfect recovery cannot be guaranteed. The description of the FEC algorithms is outside the scope of this paper but can be found in [28] Recall, however, that common FEC schemes based on Reed Solomon codes or X OR functions can ....

N. Shacham and P. McKenney, "Packet recovery in high-speed networks using coding and buffer management," in Proc. IEEE INFOCOM Conf., vol. 1, Los Alamitos, CA, 1990, pp. 124--131.

.... by Redundancy and Interleaving: An Analytical Study Alain Jean Marie, Parijat Dube, Damien Artiges, Eitan Altman Abstract This paper studies a forward error correction (FEC) scheme that reduces loss probabilities of messages, based on adding redundant packets and interleaving, as proposed in [17]. We rst show that when standard redundancy schemes are used, losses of messages occur due to the following locality phenomenon. If a packet is lost at time t, then the probability to have another loss closely after t may be signi cantly larger than the probability of a loss at a time quite ....

....locality phenomenon. If a packet is lost at time t, then the probability to have another loss closely after t may be signi cantly larger than the probability of a loss at a time quite larger than t. Losses thus have a local characteristic, and they tend to cluster. We then study the scheme of [17] that involves transmission of packets from di erent messages in a round robin way (interleaving) that does not su er from the locality of losses. We obtain expressions for loss probabilities under this scheme using di erent approaches: Ballot theorems, recursions and a purely algebraic ....

[Article contains additional citation context not shown here]

N. Shacham and P. McKenney, \Packet Recovery in High-Speed Networks Using Coding and Buer Management", Proceeding of the IEEE INFOCOM' 90, pp. 124-131, 1990.

....data. Index Terms ABL, forward error correction, Gilbert model, multimedia streaming, PLR. I. INTRODUCTION F ORWARD ERROR correction (FEC) techniques are the preferred error control schemes for multicast or interactive streaming applications. In this case, packet level FEC schemes [1] [2] provide an efficient way to fight against losses, although the perfect recovery cannot be guaranteed. Several studies have been performed to compute the FEC efficiency or the probability for data to be recovered in case of loss [3] 5] However, this parameter does not bring enough information ....

N. Shacham and P. McKenney, "Packet recovery in high-speed networks using coding and buffer management," in Proc. IEEE INFOCOM Conference, vol. 1, Los Alamitos, CA, 1990, pp. 124--131.

....and Sequence Number Checks) Correct ion (RS and XOR Codes) M 1 Figure 1 A Virtual Control Block Matrix III. THE FEC AWARE PARTIAL VIDEO SLICE DISCARD Various video packet discarding techniques have been proposed to minimize user perceived quality degradation during network congestion [8]. In this section, we propose enhancement to the Tail Drop mechanism [9] to take into account Forward Error Correction feature. The new scheme, is named FEC aware Partial video Slice Discard (FEC PSD) and performs at video slice level. Our approach is to reduce the number of corrupted slices by ....

....the buffer space of the dropped slice to a higher priority slice. The proposed approach avoids congestion increase while maintaining the mean cell transfer delay in acceptable value. This proactive strategy is performed gradually by including high priority cells if necessary. As evaluated in [8], the proposed approach can significantly improve the network performance by minimizing the transmission of non useful video data before buffer overflow. The proposed selective and adaptive partial video slice discard algorithm is highlighted bellow. FEC PSD Operation Modes and Fairness FEC PSD ....

N. Shacham et P. McKenny, "Packet Recovery in High Speed Networks using coding", in IEEE INFOCOM '90, San Francisco, CA, pp.124-131, June 1990.

....to enable most of the delayed packets to arrive before their playout time. Playout mechanisms have been investigated in [Schu92] Ramj94] and [Schu95] 2.1. 3 Data Corruption Thanks to today s technologies, a very low bit error rate can be achieved (approximately 10 14 with optical networks [Shac90]) Thus, the impact of data corruption is very small and can be neglected in speech transmissions over the Internet (except for wireless network scenarios that are not addressed in this work) 2.1.4 Packet Duplication Due to routing problems, a single packet sent by the sender can be duplicated ....

N. Shacham and P. McKenney. Packet Recovery in High-Speed Networks Using Coding and Buffer Management. Proceedings ACM SIGCOMM '90, pages 124-131, San Francisco, CA, June 1990.

....is due to the assumption of quasi stationarity for speech. This is only valid for a time period typically equivalent to one or two packets. Given these constraints, concealment and forward error recovery approaches become much less efficient as the loss burstiness increases, as shown e.g. in [9], 10] The previous arguments underline the importance of mapping application requirements with regard to their ADU format and end to end quality enhancement capa1 p 01 00 p p 10 11 p 0 1 Fig. 1. Gilbert Model bilities to network mechanisms effectively controlling the distribution of ....

N. Shacham and P. McKenney, "Packet recovery in high-speed networks using coding and buffer management," in Proceedings ACM SIGCOMM '90, San Francisco, CA, June 1990, pp. 124--131.

....modern communication media, the assumption is made that decoding is mainly impeded by packet loss. Either the packet is present and correct or it is lost. These losses are mainly caused by network congestion and the resultant bu er over ow and queuing delay. In this case, packet level FEC schemes [13 16] provide an ecient way to ght against losses, although the perfect recovery cannot be guaranteed. The details of the FEC algorithms are outside the scope of this paper but can be found in [17] Recall however that common FEC schemes based on Reed Solomon codes or X OR functions can generally ....

Shacham N. and McKenney P., \Packet recovery in High-Speed Networks Using Coding and Buer Management, " in Proceedings of the IEEE INFOCOM Conference, Los Alamitos, CA, 1990, vol. 1, pp. 124-131.

....communication media, the assumption is made that decoding is mainly impeded by packet loss. Either the packet is present and correct or it is lost. These losses are mainly caused by network congestion and the resultant buffer overflow and queuing delay. In this case, packet level FEC schemes [4, 5] provide an efficient way to fight against losses, although the perfect recovery cannot be guaranteed. Assume every block of k video packets is protected by (n k) FEC packets. If at least k out of n packets are correctly received, the underlying video information can be correctly decoded. ....

Shacham N. and McKenney P., "Packet recovery in HighSpeed Networks Using Coding and Buffer Management," in Proceedings of the IEEE INFOCOM Conference, Los Alamitos, CA, 1990, vol. 1, pp. 124--131.

....communication media, the assumption is made that decoding is mainly impeded by packet loss. Either the packet is present and correct or it is lost. These losses are mainly caused by network congestion and the resultant buffer overflow and queuing delay. In this case, packet level FEC schemes [4, 5] provide an efficient way to fight against losses, although the perfect recovery cannot be guaranteed. Assume every block of # video packets is protected by ##### FEC packets. If at least # out of # packets are correctly received, the underlying video information can be correctly decoded. ....

Shacham N. and McKenney P., "Packet recovery in HighSpeed Networks Using Coding and Buffer Management," in Proceedings of the IEEE INFOCOM Conference, Los Alamitos, CA, 1990, vol. 1, pp. 124--131.

....should require little processing and even be amenable to hardware implementation. Both coding and decoding should add a minimal amount of delay to the packets, and should require only small data storage for operation. Based on the above considerations, the scheme proposed by Shacham and McKenney ([4]) groups data packet into blocks of predetermined size, and adds to each block a number of parity packets, and their construction determines the maximum number of data packets that can be recovered. However, any subset of missing packets can be recovered by using the parity packets and the other ....

Shacham, N. and McKenney, P., "Packet Recovery in High Speed Networks Using Coding and Buffer Management," Proc. IEEE Infocom, pp. 124--131, May 1990.

....video services has simultaneously evolved in two directions. Following the concept of adaptive application, the first direction focuses on optimizing the end systems (i.e. applications and terminals) to adapt them to the impairments experienced by the packet stream in the network (e.g. 6] 7] [8]) The second direction aims to provide integrated service networks with appropriate resource management mechanisms that are able to handle multiple classes of traffic and offer different qualities of service. The ATM Transfer Capacities [9] and the IETF Internet Service Model [10] 11] 12] are ....

....by using the two adjacent frames. Concealment mechanisms for missing voice packets are described in [19] and [20] ffl Forward Error Correction mechanisms consist in adding redundancy information by the transmitter. There are several FEC mechanisms proposed in the literature (e.g. 21] and [8]) The potential of FEC mechanisms to recover from losses widely depends on the packet loss process. For a given packet loss ratio, a FEC mechanism is more effective when lost packets are dispersed throughout the packet stream [7] Unlike other communication services where the delay average and ....

N. Shacham and P. McKenney, "Packet Recovery in High-Speed Networks using Coding and Buffer Management", in Proceedings of INFOCOM'90. IEEE, May 1990.

....are much cheaper to decode when only a few erasures are expected) In some cases, erasure codes are trivial to build. As an example, an (n; 1) code is simply implemented by retransmitting n times the only source packet, and a (k 1; k) code can be implemented by XOR ing the k source packets [9]. For generic n and k, erasure codes can be constructed and studied basing on the properties of linear algebra over finite fields. A complete discussion of the subject can be found in [1] while [5, 8] give a simplified description more closely related to erasure codes and networking protocols. 3 ....

N.Shacham, P.McKenney, "Packet recovery in high-speed networks using coding and buffer management", Proc. IEEE Infocom'90, San Francisco, CA, pp.124-131, May 1990

....time critical, e.g. resource reservation for connection setup. The discussion here mainly applies to dynamic control mechanisms. 12 Finally, the fourth solution is to take preventive action to minimize any reliance on feedback. For error control, this means using forward error correction [Shacham90, Biersack92] to anticipate errors and provide information redundancy, allowing the receivers to reconstruct the information without asking the sender to retransmit. For flow control, this means reserving resources so that receivers and intermediate switching nodes are always able to support the flow rate ....

N. Shacham and P. McKenney, "Packet Recovery in High-Speed Networks Using Coding and Buffer Management," Proc. IEEE INFOCOM '90, pp. 124-131 (June 1990).

....on packet boundaries (burst erasures) and thus burst erasure correcting codes are the appropriate form of forward error correction for these environments. The effectiveness of burst erasure codes is strongly dependent on the burstiness of the network loss process and the degree of overcoding. In [52] a block code for cell switched networks based on the exclusive OR operation was used to generate one or two redundant cells in a block of k cells. The study concluded that the coverage from this code was not very useful and suggested the need for sophistication in multiplexing and or cell ....

N. Shacham and P. McKenny. Packet Recovery in High-Speed Networks using Coding. INFOCOM 1990, pages 124--131, June 1990. 101

....you could you refer to it as a piggybacking FEC scheme. We decided to call it piggyback protected transmission . Many FEC mechanisms proposed in the literature involve exclusive OR operations. The idea is to send every th packet a redundant packet obtained by exclusive ORing the other packets [39]. This mechanism can recover from a single loss within a series of packets, but it increases the sending rate of the source by a factor of , and it adds latency since packets have to be received before the lost packet can be reconstructed. The approach developed by the MICE researchers is a rather ....

N. Shacham and P. McKenney, `Packet recovery in high-speed networks using coding and buffer management', Proceedings IEEE Infocom 90 in San Francisco, pp.124-131, May 1990

.... such as real time MPEG video over high speed wide area networks, packet level FEC has generated renewed interest due to the ARQ mechanism s i.e. retransmission of lost messages inherent inadequacy at handling timing constraints when subject to packet drops and large end to end network latencies [4,6,9,39,51,54]. That is, by the time a lost or damaged MPEG frame is detected and retransmitted, its arrival at the receiver is too late to be useful. Due to the low bit error rates associated with modern communication media, the assumption is made that decoding is mainly impeded by packet loss i.e. ....

....in such situations with respect to achieving higher throughput. Only recently, however, have studies been carried out for packet level FEC which possess additional structure due to the fact that dependencies can stem from queueing e ects as well as the characteristics of the trac source itself. In [54] simulation is used to show that burstiness can lead to recovery properties signi cantly di erent from those that stem from assuming independence. This has been con rmed and further analyzed by Cidon et al. 15,16] albeit excluding LRD input processes. In most recent work [57] we have extended ....

N. Shacham and P. McKenny. Packet recovery in high-speed networks using coding. In Proc. IEEE INFOCOM '90, pages 124-131, 1990.

....in the receiver and the sender does not need to know which packets were lost. Since no retransmission is required, no time penalty is imposed by the loss recovery. For this reason, FEC is often considered to be a preferred error recovery method for real time interactive media communication [13][14] 15] The drawback of FEC is the increased bandwidth required by the redundant data. Furthermore, its ability to recover lost packets is strictly dependent on the loss characteristics of the underlying network. Specifically, FEC cannot recover from a large burst of packet loss, which we ....

N.Shacham, P. McKenney. Packet recovery in high-speed networks using coding and buffer management. Proc. IEEE INFOCOM, Vol.1, San Francisco, CA, June,1990.

....but can not tolerate any loss. To support these applications, ARQ is more suitable than FEC, because ARQ makes a more efficient use of the bandwidth by only retransmitting the messages that are lost. 5 Related Work Loss recovery using FEC has been studied previously [KITA 90, OHTA 90, OHTA 91, SHAC 90, ZHAN 91] Our work differentiates itself in several ways. 1) Our traffic model for the video sources is derived from an actual movie while previous models assumed that the inter arrival times of video data are exponentially or hyper exponentially distributed. 2) The code used to generate the ....

N. Shacham and P. McKenny, "Packet Recovery in High-Speed Networks using Coding", Proc. INFOCOM 90, pp. 124--131, San Francisco, CA, June 1990.

No context found.

N. Shacham and P. McKenny, "Packet Recovery in High-Speed Networks using Coding and Buffer Management ", Proceedings of IEEE INFOCOM '90, San Francisco, CA, June 1990.

No context found.

N. Shacham and P. McKenny, "Packet Recovery in HighSpeed Networks using Coding and Buffer Management", Proceedings of IEEE INFOCOM '90, San Francisco, CA, June 1990.

No context found.

N. Shacham and P. McKenny, "Packet Recovery in High-Speed Networks using Coding and Buffer Management ", Proceedings of IEEE INFOCOM '90, San Francisco, CA, June 1990.

No context found.

N. Shacham and P. McKenny, "Packet Recovery in HighSpeed Networks using Coding and Buffer Management", Proceedings of IEEE INFOCOM '90, San Francisco, CA, June 1990.

No context found.

N. Shacham and P. McKenney, "Packet recovery in high-speed networks using coding and buffer management," in Proc. INFOCOM 90, vol. 1, San Francisco, CA, USA, June 1990, pp. 124--131.

No context found.

N. Shacham and P. McKenny, "Packet Recovery in High-Speed Networks using Coding and Buffer Management ", Proceedings of IEEE INFOCOM '90, San Francisco, CA, June 1990.

No context found.

N. Shacham and P. McKenney. Packet recovery in high-speed networks using coding and buer management. In Proc. INFOCOM 90, volume 1, pages 124131, San Francisco, CA, USA, June 1990.

No context found.

N. Shacham and P. McKenney. Packet Recovery in High-Speed Networks using Coding and Buer Management. In Proc. INFOCOM 90, volume 1, pages 124131, San Francisco, CA, USA, June 1990.

No context found.

N. Shacham and P. McKenney, "Packet Recovery in High-Speed Networks Using Coding and Bu#er Management", IEEE INFOCOM, Jun. 1990.

No context found.

N. Shacham, P. McKenney, Packet recovery in high-speed networks using coding and buffer management, in: Proceedings of INFOCOM '90, San Francisco, CA, 1990.

No context found.

N. Shacham, P. McKenney, Packet recovery in high-speed networks using coding and buffer management, in: INFOCOM'90, San Francisco, CA, 1990, pp. 124--131.

No context found.

Shacham N. and McKenney P., Packet recovery in High-Speed Networks Using Coding and Buer Management, in Proceedings of the IEEE INFOCOM Conference, vol. 1, (Los Alamitos, CA), pp. 124131, 1990.

No context found.

N. Shacham and P. McKenny, "Packet Recovery in HighSpeed Networks using Coding and Buffer Management", Proceedings of IEEE INFOCOM '90, San Francisco, CA, June 1990.

No context found.

N. Shacham and P. McKenny, "Packet Recovery in HighSpeed Networks using Coding and Buffer Management", Proceedings of IEEE INFOCOM '90, San Francisco, CA, June 1990.

No context found.

N. Shacham and P. McKenny, "Packet Recovery in High-Speed Networks using Coding and Buffer Management ", Proceedings of IEEE INFOCOM '90, San Francisco, CA, June 1990.

No context found.

N. Shacham and P. McKenny, "Packet Recovery in High-Speed Networks using Coding and Buffer Management ", Proceedings of IEEE INFOCOM '90, San Francisco, CA, June 1990.

No context found.

N. Shacham & P. McKenney, "Packet Recovery in HighSpeed Networks Using Coding and Buffer Management," Proc. IEEE Infocom '90, vol. 1, pp. 124-31, San Francisco, CA, June 1990.

No context found.

N. Shacham and P. McKenney,#Packet recovery in high-speed networks using coding and bu#er management," in Proceedings ACM SIGCOMM '90, pp. 124#131, #San Francisco, CA#, June 1990.

No context found.

N. Shacham and P. McKenney,#Packet recovery in high-speed networks using coding and bu#er management," in Proceedings ACM SIGCOMM '90, pp. 124#131, #San Francisco, CA#, June 1990.

No context found.

N. Shacham and P. McKenney, "Packet recovery in high-speed networks using coding and buffer management," in Proceedings ACM SIGCOMM '90, pp. 124--131, (San Francisco, CA), June 1990.

No context found.

Shacham N. and McKenney P., \Packet recovery in High-Speed Networks Using Coding and Buer Management, " in Proceedings of the IEEE INFOCOM Conference, vol. 1, pp. 124-131, (Los Alamitos, CA), 1990.

First 50 documents  Next 50

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