Packet loss

About: Packet loss is a research topic. Over the lifetime, 21235 publications have been published within this topic receiving 302453 citations.

Measuring link bandwidths using a deterministic model of packet delay

Abstract: We describe a deterministic model of packet delay and use it to derive both the packet pair [2] property of FIFO-queueing networks and a new technique packet tailgating) for actively measuring link bandwidths. Compared to previously known techniques, packet tailgating usually consumes less network bandwidth, does not rely on consistent behavior of routers handling ICMP packets, and does not rely on timely delivery of acknowledgments.

On Packet Switches with Infinite Storage

Abstract: Most prior work on congestion in datagram systems focuses on buffer management. We find it illuminating to consider the case of a packet switch with infinite storage. Such a packet switch can never run out of buffers. It can, however, still become congested. The meaning of congestion in an infinite-storage system is explored. We demonstrate the unexpected result that a datagram network with infinite storage, first-in, first-out queueing, at least two packet switches, and a finite packet lifetime will, under overload, drop all packets. By attacking the problem of congestion for the infinite-storage case, we discover new solutions applicable to switches with finite storage.

Characterizing End-to-End Packet Delay and Loss in the Internet

Abstract: We use the measured round trip delays of small UDP probe packets sent at regular time intervals to characterize the end-to-end packet delay and loss behavior in the Internet. By varying the interval between probe packets, it is possible to study the structure of the Internet load over different time scales. In this paper, the time scales of interest range from a few milliseconds to a few minutes. Our observations agree with results obtained by others using simulation and experimental approaches. For example, our estimates of Internet workload are consistent with the hypothesis of a mix of bulk traffic with larger packet size, and interactive traffic with smaller packet size. The interarrival time distribution for Internet packets is consistent with an exponential distribution. We also observe a phenomenon of compression (or clustering) of the probe packets similar to the acknowledgement compression phenomenon recently observed in TCP. Our results also show interesting and less expected behavior. For example, we find that the losses of probe packets are essentially random when the probe traffic uses a small fraction of the available bandwidth.

A multichannel CSMA MAC protocol for multihop wireless networks

Abstract: We describe a new carrier-sense multiple access (CSMA) protocol for multihop wireless networks, sometimes also called ad hoc networks. The CSMA protocol divides the available bandwidth into several channels and selects an idle channel randomly for packet transmission. It also employs a notion of "soft" channel reservation as it gives preference to the channel that was used for the last successful transmission. We show via simulations that this multichannel CSMA protocol provides a higher throughput compared to its single channel counterpart by reducing the packet loss due to collisions. We also show that the use of channel reservation provides better performance than multichannel CSMA with purely random idle channel selection.

JAM: a jammed-area mapping service for sensor networks

Abstract: Preventing denial-of-service attacks in wireless sensor networks is difficult primarily because of the limited resources available to network nodes and the ease with which attacks are perpetrated Rather than jeopardize design requirements which call for simple, inexpensive, mass-producible devices, we propose a coping strategy that detects and maps jammed regions We describe a mapping protocol for nodes that surround a jammer which allows network applications to reason about the region as an entity, rather than as a collection of broken links and congested nodes This solution is enabled by a set of design principles: loose group semantics, eager eavesdropping, supremacy of local information, robustness to packet loss and failure, and early use of results Performance results show that regions can be mapped in 1-5 seconds, fast enough for real-time response With a moderately connected network, the protocol is robust to failure rates as high as 25 percent