ATM交换机EPD性能分析

利用ＡＴＭ网络传送ＴＣＰ分组时的一个重要问题是分组中任何一个信元丢失都半导致分组的重传,为解决这一问题,一个方法是发生拥塞时交换机有选择丢弃信元,在有限的存储器容量下,交换机智能化的信元集合源状态与缓存器占用的联合分布,基于这一结果,本文推导出的ＥＰＤ系统的分组丢失率的上界和下界,通过分析比较发现,使用ＥＰＤ的系统,当门限设为缓存器大小时其性能将优于无控制系统。     

A fast parallel-tree switch architecture for ATM networks

In this paper we present a novel fast packet switch architecture based on Banyan interconnection networks, called parallel-tree Banyan switch fabric (PTBSF). It consists of parallel Banyans (multiple outlets) arranged in a tree topology. The packets enter at the topmost Banyan. Internal conflicts are eliminated by using a conflict-free 3 × 4 switching element which distributes conflicting cells over different Banyans. Thus, cell loss may occur only at the lowest Banyan. Increasing the number of Banyans leads to a noticeable decrease in cell loss rate. The switch can be engineered to provide arbitrarily high throughput and low cell loss rate without the use of input buffering or cell pre-processing. The performance of the switch is evaluated analytically under uniform traffic load and by simulation, under a variety of asynchronous transfer mode (ATM) traffic loads. Compared to other proposed architectures, the switch exhibited stable and excellent performance with respect to cell loss and switching delay for all studied conditions as required by ATM traffic sources. The advantages of PTBSF are modularity, regularity, self-routing, low processing overhead, high throughput and robustness, under a variety of ATM traffic conditions. © 1998 John Wiley & Sons, Ltd.     

一种自适应早期包丢弃方案

该文提出了一种自适应早期包丢弃方案,与以ATM网络支持的TCP应用中通常采用的早期包丢弃方案相比,该方案能减少ATM网络交换结点所需的缓存数量,提高ATM网络中IP流量的完好流量通过率,并降低丢包率。     

Guaranteed frame rate: a better service for TCP/IP in ATM networks

Guaranteed frame rate, approved by the ATM Forum, is expected to become an important service category to efficiently support TCP/IP traffic in ATM networks. We first describe the GFR traffic contract in detail. We then present different types of switch implementations that have been proposed to support GFR. We analyze the performance of three of these switch implementations by simulations in two different network environments. These simulations show that the scheduler-based implementations provide a much better performance than the simple switch implementation. However, we also show that coupling an active packet discard mechanism to a scheduler-based switch implementation does not produce a performance gain when many TCP connections are multiplexed inside one ATM VC     

A new ATM adaptation layer for TCP/IP over wireless ATM networks

This paper describes the design and performance of a new ATM adaptation layer protocol (AAL‐T) for improving TCP performance over wireless ATM networks. The wireless links are characterized by higher error rates and burstier error patterns in comparison with the fiber links for which ATM was introduced in the beginning. Since the low performance of TCP over wireless ATM networks is mainly due to the fact that TCP always responds to all packet losses by congestion control, the key idea in the design is to push the error control portion of TCP to the AAL layer so that TCP is only responsible for congestion control. The AAL‐T is based on a novel and reliable ARQ mechanism to support quality‐critical TCP traffic over wireless ATM networks. The proposed AAL protocol has been validated using the OPNET tool with the simulated wireless ATM network. The simulation results show that the AAL‐T provides higher throughput for TCP over wireless ATM networks compared to the existing approach of TCP with AAL 5. This revised version was published online in July 2006 with corrections to the Cover Date.     

Packet loss performance of selective cell discard schemes in ATMswitches

In transport-layer protocols such as TCP over ATM networks, a packet is discarded when one or more cells of that packet are lost, and the destination node then requires its source to retransmit the corrupted packet. Therefore, once one of the cells constituting a packet is lost, its subsequent cells of the corrupted packet waste network resources. Thus, discarding those cells will enable us to efficiently utilize network resources, and will improve the packet loss probability. We focus on tail dropping (TD) and early packet discard (EPD) as selective cell discard schemes which enforce the switches to discard some of the arriving cells instead of relaying them. We exactly analyze the packet loss probability in a system applying these schemes. Their advantages and limits are then discussed based on numerical results derived through the analysis     

An efficient ATM network switch scheduling

The problem of allocating network resources to application sessions backlogged at an individual switch has a great impact on the end-to-end delay and throughput guarantees offered by the network. There exists a class of algorithms based on weighted fair queueing (WFQ) for scheduling packets which are work-conserving and they guarantee fairness to the backlogged sessions. These algorithms also apply to ATM networks with a packet equal to a single cell or an ATM block (of fixed size). Bursts are groups of varying numbers of cells. We generalize WFQ to schedule bursts. Our motivation is to derive an adaptive algorithm which generalizes the (fixed size) packet level to a varying size packet level. The new algorithm enhances the performance of the switch service for many important applications. The proposed scheme maintains the work-conserving property, and also provides throughput and fairness guarantees. The worst-case delay bound is also given. We use simulation to study the performance characteristics of our algorithm. Our results demonstrate the efficiency of the new algorithm.     

Analysis of packet loss in transport layer over ATM networks

In asynchronous transfer mode (ATM) networks, when cells are lost due to congestion, packets containing the lost cells should be retransmitted in the transport layer, which manages the end-to-end communication. The probability that a packet contains at least one lost cell depends on the packet length. It is thus very likely that the performance of the end-to-end communication is influenced by the packet length. In this paper, we analyse packet loss probability and the achievable maximum throughput when a block of data is divided into packets of fixed size and the lost packets are retransmitted based on the selective repeat automatic repeat request (ARQ). Through this analysis, we examine the effect of packet length and peak cell transmission rate on the performance measures mentioned above. © 1997 John Wiley & Sons, Ltd.     

Explicit window adaptation: a method to enhance TCP performance

We study the performance of TCP in an internetwork consisting of both rate-controlled and nonrate-controlled segments. A common example of such an environment occurs when the end systems are part of IP datagram networks interconnected by a rate-controlled segment, such as an ATM network using the available bit rate (ABR) service. In the absence of congestive losses in either segment, TCP keeps increasing its window to its maximum size. Mismatch between the TCP window and the bandwidth-delay product of the network results in accumulation of large queues and possibly buffer overflows in the devices at the edges of the rate-controlled segment, causing degraded throughput and unfairness. We develop an explicit feedback scheme, called explicit window adaptation, based on modifying the receiver's advertised window in TCP acknowledgments returning to the source. The window size indicated to TCP is a function of the free buffer in the edge device. Results from simulations with a wide range of traffic scenarios show that this explicit window adaptation scheme can control the buffer occupancy efficiently at the edge device, and results in significant improvements in packet loss rate, fairness, and throughput over a packet discard policy such as random early detection (RED)     

TCP-Jersey for wireless IP communications

Improving the performance of the transmission control protocol (TCP) in wireless Internet protocol (IP) communications has been an active research area. The performance degradation of TCP in wireless and wired-wireless hybrid networks is mainly due to its lack of the ability to differentiate the packet losses caused by network congestions from the losses caused by wireless link errors. In this paper, we propose a new TCP scheme, called TCP-Jersey, which is capable of distinguishing the wireless packet losses from the congestion packet losses, and reacting accordingly. TCP-Jersey consists of two key components, the available bandwidth estimation (ABE) algorithm and the congestion warning (CW) router configuration. ABE is a TCP sender side addition that continuously estimates the bandwidth available to the connection and guides the sender to adjust its transmission rate when the network becomes congested. CW is a configuration of network routers such that routers alert end stations by marking all packets when there is a sign of an incipient congestion. The marking of packets by the CW configured routers helps the sender of the TCP connection to effectively differentiate packet losses caused by network congestion from those caused by wireless link errors. This paper describes the design of TCP-Jersey, and presents results from experiments using the NS-2 network simulator. Results from simulations show that in a congestion free network with 1% of random wireless packet loss rate, TCP-Jersey achieves 17% and 85% improvements in goodput over TCP-Westwood and TCP-Reno, respectively; in a congested network where TCP flow competes with VoIP flows, with 1% of random wireless packet loss rate, TCP-Jersey achieves 9% and 76% improvements in goodput over TCP-Westwood and TCP-Reno, respectively. Our experiments of multiple TCP flows show that TCP-Jersey maintains the fair and friendly behavior with respect to other TCP flows.     

Design of a non-blocking shared-memory copy network for ATM

The design of a copy network is presented for use in an ATM (asynchronous transfer mode) switch supporting BISDN (broadband integrated services digital network) traffic. Inherent traffic characteristics of BISDN services require ATM switches to handle bursty traffic with multicast connections. In typical ATM switch designs a copy network is used to replicate multicast cells before being forwarded to a point-to-point routeing network. In such designs, a single multicast cell enters the switch and is replicated once for each multicast connection. Each copy is forwarded to the routeing network with a unique destination address and is routed to the appropriate output port. Non-blocking copy networks permit multiple cells to be multicasted at once, up to the number of outputs of the copy network. Another critical feature of ATM switch design is the location of buffers for the temporary storage of transmitted cells. Buffering is required when multiple cells require a common switch resource for transmission. Typically, one cell is granted the resource and is transmitted while the remaining cells are buffered. Current switch designs associate discrete buffers with individual switch resources. Discrete buffering is not efficient for bursty traffic as traffic bursts can overflow individual switch buffers and result in dropped cells, while other buffers are under-used. A new non-blocking copy network is presented in this paper with a shared-memory input buffer. Blocked cells from any switch input are stored in a single shared input buffer. The copy network consists of three banyan networks and shared-memory queues. The design is scalable for large numbers of inputs due to low hardware complexity, O (N log₂ N), and distributed operation and control. It is shown in a simulation study that a switch incorporating the shared-memory copy network has increased throughput and lower buffer requirements to maintain low packet loss probability when compared to a switch with a discrete buffer copy network.     

Prevention of deadlocks and livelocks in lossless backpressured packet networks

No packets will be dropped inside a packet network, even when congestion builds up, if congested nodes send backpressure feedback to neighboring nodes, informing them of unavailability of buffering capacity-stopping them from forwarding more packets until enough buffer becomes available. While there are potential advantages in backpressured networks that do not allow packet dropping, such networks are susceptible to a condition known as deadlock in which throughput of the network or part of the network goes to zero (i.e., no packets are transmitted). In this paper, we describe a simple, lossless method of preventing deadlocks and livelocks in backpressured packet networks. In contrast with prior approaches, our proposed technique does not introduce any packet losses, does not corrupt packet sequence, and does not require any changes to packet headers. It represents a new networking paradigm in which internal network losses are avoided (thereby simplifying the design of other network protocols) and internal network delays are bounded.     

Enhancing Fairness and Throughput of TCP in Heterogeneous Wireless Access Networks

Most of the recent research on TCP over heterogeneous wireless networks has concentrated on differentiating between packet drops caused by congestion and link errors, to avoid significant throughput degradations due to the TCP sending window being frequently shut down, in response to packet losses caused not by congestion but by transmission errors over wireless links. However, TCP also exhibits inherent unfairness toward connections with long round-trip times or traversing multiple congested routers. This problem is aggravated by the difference of bit-error rates between wired and wireless links in heterogeneous wireless networks. In this paper, we apply the TCP Bandwidth Allocation (TBA) algorithm, which we have proposed previously, to improve TCP fairness over heterogeneous wireless networks with combined wireless and wireline links. To inform the sender when congestion occurs, we propose to apply Wireless Explicit Congestion Notification (WECN). By controlling the TCP window behavior with TBA and WECN, congestion control and error-loss recovery are effectively separated. Further enhancement is also incorporated to smooth traffic bursts. Simulation results show that not only can the combined TBA and WECN mechanism improve TCP fairness, but it can maintain good throughput performance in the presence of wireless losses as well. A salient feature of TBA is that its main functions are implemented in the access node, thus simplifying the sender-side implementation.     

Improving Reno and New-Reno's performances over OBS networks without SACK

In TCP over optical burst switching (OBS) networks, consecutive multiple packet losses are common since an optical burst usually contains a number of consecutive packets from the same TCP sender. It has been proved that over OBS networks Reno and New-Reno achieve lower throughput performances than that of SACK, which can address the inefficiency of Reno and New-Reno in dealing with consecutive multiple packet losses. However, SACK adopts complex mechanisms not only at the sender's but also at the receiver's protocol stack, and thus has a higher difficulty in deployment.In this paper we propose B-Reno, a new TCP implementation designed for TCP over OBS networks. Using some simple modifications to New-Reno only at the sender's protocol stack, B-Reno can overcome the inefficiencies of Reno and New-Reno in dealing with consecutive multiple packet losses and thus improve their throughputs over OBS networks. Moreover, B-Reno can also achieve performance similar with that of SACK over OBS networks while avoiding SACK's difficulty in deployment due to complex mechanisms at both the sender's and the receiver's protocol stack.     

The dual flow control problem of TCP over ATM ABR services

In this paper, we investigate the dual control problem—TCP flow control at the TCP layer and ABR flow control at the ATM layer. First, we observe that TCP flow control and ABR flow control cannot co‐operate well. The worst case is that the slow start after packet loss causes high but unused ACR (Allowed Cell Rate) which raises the potential of cell loss and an underflowed switch queue which reduces ABR throughput. We suggest to implement a use‐it‐or‐lose‐it policy for ABR and fast recovery for TCP to avoid these phenomena. Copyright © 1999 John Wiley & Sons, Ltd.     

A growable packet (ATM) switch architecture: design principles andapplication

The problem of designing a large high-performance, broadband packet of ATM (asynchronous transfer mode) switch is discussed. Ways to construct arbitrarily large switches out of modest-size packet switches without sacrificing overall delay/throughput performance are presented. A growable switch architecture is presented that is based on three key principles: a generalized knockout principle exploits the statistical behaviour of packet arrivals and thereby reduces the interconnect complexity, output queuing yields the best possible delay/throughput performance, and distributed intelligence in routing packets through the interconnect fabric eliminates internal path conflicts. Features of the architecture include the guarantee of first-in-first-out packet sequence, broadcast and multicast capabilities, and compatibility with variable-length packets, which avoids the need for packet-size standardization. As a broadband ISDN example, a 2048×2048 configuration with building blocks of 42×16 packet switch modules and 128×128 interconnect modules, both of which fall within existing hardware capabilities, is presented     

Packet network simulation: speedup and accuracy versus timinggranularity

This paper describes a new technique that can speedup simulation of high-speed, wide-area packet networks by one to two orders of magnitude. Speedup is achieved by coarsening the representation of network traffic from packet-by-packet to train-by-train, where a train represents a cluster of closely spaced packets. Coarsening the timing granularity creates longer trains and makes the simulation proceed more quickly since the cost of processing trains is independent of train size. Coarsening the timing granularity introduces, of course, a degree of approximation. This paper presents experiments that evaluate our coarse time-grain simulation technique for first in/first out (FIFO) switched, TCP/IP, and asynchronous transfer mode (ATM) networks carrying a mix of data and streaming traffic. We show that delay, throughput, and loss rate can frequently be estimated within a few percent via coarse time-grain simulation. This paper also describes how to apply coarse time-grain simulation to other switch disciplines. Finally, this paper introduces three more simulation techniques which together can double the performance of well written packet simulators without trading with the simulation accuracy. These techniques reduce the number of outstanding simulation events and reduce the cost of manipulating the event list     

A high-capacity broadband packet switch architecture based on a multilink approach

Broadband packet networks based on asynchronous transfer mode (ATM) are expected to provide a wide range of services, including motion video, voice, data and image. When these networks become prevalent, some applications such as motion video and high-speed LAN interconnections will place a very large bit rate requirement on the channels. Currently, the physical layer supported by the synchronous optical network (SONET) allows the transmission of up to 2.4 Gbit/s with the OC-48 optical interface. However, it is not feasible for the electronic packet switch to route packets at this rate on a single link. In this paper we present a design of a broadband packet switch that uses multiple links in parallel to realize a high-speed channel. This implementation permits the switch to operate at the lower link rate, which can be at 150 Mbit/s, while having the ability to support a virtual circuit at a higher rate (up to 2.4 Gbit/s). The main contribution of the design is that packet sequence on a channel is still maintained even though packets are allowed to use any of the links belonging to the same channel. Besides allowing the switch to function at a slower rate than the transmission channel rate, the implementation of the multilinks benefits from statistical multiplexing gain. Analytical results show the performance advantages of multilink design with respect to delay, throughput and packet loss probability.     

Impact of mobility on TCP/IP: an integrated performance study

This paper presents a simulation analysis of the impact of mobility on TCP/IP augmented with features to support host mobility in wide area networks. Our results show that the existing version of TCP can yield low throughput in highly mobile environments due to the fact that TCP cannot discriminate packets dropped due to hand-offs with those dropped due to congestion in one or more network resources. As a result, TCP invokes a congestion recovery process when packets are lost during internetwork hand-offs of the mobile host. We investigate a proposal in which the transport layer explicitly receives information from the network layer of any ongoing mobility. We show that by effectively capitalizing this information, TCP can appropriately extend the slow-start phase in the recovery process and achieve higher throughput. Based on the simulation analysis we also show the robustness of this scheme in the presence of both host mobility and network congestion     

RED分组丢弃算法性能研究

 本文研究了在ATM交换机上实现的RED算法的性能.在固定有效带宽、时变有效带宽情况下和同种、异种业务环境下,研究了RED算法的通过率、公平性和时延等性能.经研究表明:RED算法有必要与EPD算法相结合,构成RED+EPD算法.采用RED+EPD算法的ATM交换机通过控制平均排队长度,有效地减小了交换机的平均排队时延.通过与其他分组丢弃算法进行性能比较表明:采用RED+EPD算法的ATM交换机,可提供比EPD算法略高的通过率,更好的公平性和更低的排队时延,能较好地支持具有时延要求的业务.