Explicit allocation of best-effort packet delivery service

Abstract

Abstract-This paper presents the "allocated-capacity" framework for providing different levels of best-effort service in times of network congestion. The "allocatedcapacity" framework-extensions to the Internet protocols and algorithms-can allocate bandwidth to different users in a controlled and predictable way during network congestion. The framework supports two complementary ways of controlling the bandwidth allocation: sender-based and receiver-based. In today's heterogeneous and commercial Internet the framework can serve as a basis for charging for usage and for more efficiently utilizing the network resources. We focus on algorithms for essential components of the framework: a differential dropping algorithm for network routers and a tagging algorithm for profile meters at the edge of the network for bulk-data transfers. We present simulation results to illustrate the effectiveness of the combined algorithms in controlling transmission control protocol (TCP) traffic to achieve certain targeted sending rates.Index Terms-Internet protocol, packet networks, quality of service, rate control, TCP.

Packet Drop Technique for Differentiated Services in Wired Ship Area Networks

“…또한 IntServ는 기존 인터넷의 하부구조를 수정해야 하는 단점도 있다 [8] . [9,10] . 000 120 1 020 121 1 110 224 1 001 300 3 021 220 2 111 330 2 002 301 3 022 221 2 112 331 2 003 302 3 023 311 3 113 332 2 004 222 2 024 312 3 114 333 2 010 122 1 100 223 1 200 340 1 011 123 1 101 320 2 201 341 1 012 303 3 102 321 2 202 342 1 013 304 3 103 322 2 203 343 1 014 310 3 104 323 2 204 344 1 …”

TCP-aware packet marking in networks with DiffServ support

“…Therefore, we select a set of RED/ RIO parameters and do not change them through all our analysis. In particular, in the case of RED we use the parameters suggested in [3], i.e., ðmin th ; max th ; max p Þ ¼ ð10; 30; 0:05Þ. In the case of RIO we use the same set of parameters for the OUT packets, and we use (40, 70, 0.02) for the IN packets.…”

“…In particular, we consider an AF class of service, based on the RIO (RED with In and Out) [3] buffer management in core routers. RIO distinguishes simply between two classes of packets, Inprofile (IN) and Out-profile (OUT), based on a one-bit marking in packet headers.…”

Proportional Bandwidth Distribution in IP Networks Implementing the Assured Forwarding PHB

“…The bottleneck is placed between the edge nodes E 1 and E 2 , since the sources transmit at link rate. For the AF PHB, edge nodes employ Dual Queuing [15] instead of RIO (RED (In and Out packets)) [8]. Dual Queuing (DQ) is a buffer management scheme that treats in and out packets differently but avoiding interference between them.…”

“…This algorithm allocates back out-of-profile bandwidth to in-profile bandwidth in proportion to the target rates, what presumably leads to higher assured bandwidth for flows with high target rates. On the other hand, algorithms such as TSW (Time Sliding Window) [8] or ETSW (Enhanced Time Sliding Window) [9] were employed to compare the performance of EBM [5] in terms of excess bandwidth sharing. Although neither TSW nor ETSW were thought to carry out a proportional distribution of excess bandwidth, the widespread use of TSW turned them into classical references.…”