Andrea Bianco, R. Birke, J. Finochietto, Luca Giraudo, F. Marenco, M. Mellia, Azeem J. Khan, D. Manjunath
{"title":"Control and management plane in a multi-stage software router architecture","authors":"Andrea Bianco, R. Birke, J. Finochietto, Luca Giraudo, F. Marenco, M. Mellia, Azeem J. Khan, D. Manjunath","doi":"10.1109/HSPR.2008.4734449","DOIUrl":"https://doi.org/10.1109/HSPR.2008.4734449","url":null,"abstract":"Software routers based on personal computer (PC) architectures are receiving increasing attention in the research community. However, a router based on a single PC suffers from limited bus and central processing unit (CPU) bandwidth, high memory access latency, limited scalability in terms of number of network interface cards, and lack of resilience mechanisms. Multi-stage architectures created by interconnecting several PCs are an interesting alternative since they allow to i) increase the performance of single-software routers, ii) scale router size, iii) distribute packet-forwarding and control functionalities, iv) recover from single-component failures, and v) incrementally upgrade router performance. However, a crucial issue is to hide the internal details of the interconnected architecture so that the architecture behaves externally as a single router, especially when considering the control and the management plane. In this paper, we describe a control protocol for a previously proposed multi-stage architecture based on PC interconnection. The protocol permits information exchange among internal PCs to support: i) configuration of the interconnected architecture, ii) packet forwarding, iii) routing table distribution, iv) management of the internal devices. The protocol is operating system independent, since it interacts with software routing suites such as Quagga and Xorp, and it is under test in our labs on a small-scale prototype of the multi-stage router.","PeriodicalId":130484,"journal":{"name":"2008 International Conference on High Performance Switching and Routing","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127846616","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"First-fit scheduling for multi-stage packet switching networks","authors":"J. Hui, Lingie Li","doi":"10.4304/jcm.5.3.205-210","DOIUrl":"https://doi.org/10.4304/jcm.5.3.205-210","url":null,"abstract":"A distributed algorithm for input scheduling through a single stage or 3-stage Clos network is proposed. Similar to other scheduling algorithms, full loading is achievable, while randomization of input selection is replaced by the use of arrival timing. This distributed algorithm applies to both cases of fixed versus variable length packets. We simulated the performance for both cases. A case is made that for multigigabit per second link speed, variable length packets should be used to simplify scheduling and avoid fragmentation, with acceptable increase in delay.","PeriodicalId":130484,"journal":{"name":"2008 International Conference on High Performance Switching and Routing","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122608531","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Proactive routing under conversion cascading constraint in optical burst switching networks","authors":"Xingbo Gao, M. Bassiouni, Guifang Li","doi":"10.1109/HSPR.2008.4734419","DOIUrl":"https://doi.org/10.1109/HSPR.2008.4734419","url":null,"abstract":"The negative impact of cascaded wavelength conversions has been largely ignored in optical burst switching performance evaluations. When optical bursts are transmitted all optically from source to destination, each wavelength conversion performed along the lightpath causes some signal-to-noise deterioration. If the distortion of the signal quality becomes significant enough, the receiver would not be able to recover the original data. In this paper we examine the performance degradation when an upper bound on the number of wavelength conversions that a signal can go through is enforced. We refer to this constraint as conversion cascading constraint. We propose a novel proactive routing scheme that considers the instantaneous link congestion at the moment when bursts arrive. It has three major advantages: 1) utilize the same offset times for the same node pairs while providing dynamic routing without using any fiber delay lines; 2) decrease burst loss probability to a great extent; 3) mitigate unfairness among bursts with different hop counts, which is even worse under conversion cascading constraint if without any special treatment.","PeriodicalId":130484,"journal":{"name":"2008 International Conference on High Performance Switching and Routing","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133977772","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Achieving 100% throughput in two-stage switch with feedback","authors":"Bing Hu, K. Yeung","doi":"10.1109/HSPR.2008.4734444","DOIUrl":"https://doi.org/10.1109/HSPR.2008.4734444","url":null,"abstract":"The feedback-based two-stage switch elegantly solves the packet mis-sequencing problem without sacrificing the packet delay performance. In this paper, we prove that for any arbitrary work-conserving input port scheduler, the feedback-based two-stage switch can achieve 100% throughput under a speedup of two. We first establish the fluid model for feedback-based two-stage switch. Based on it, a joint queue occupancy function is derived by taking all packets arrived at input i and all packets destined for output j into account. By showing there is no fluid accumulation for this function, we prove that the feedback-based switch is stable under a speed up of two. Compared with other switches operating at a speedup of two, feedback-based two-stage switch does not require a central scheduler as its switch fabrics are configured according to a pre-determined and periodic sequence of switch configurations. It is thus more scalable and practical for implementation even under a speedup of two.","PeriodicalId":130484,"journal":{"name":"2008 International Conference on High Performance Switching and Routing","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129169745","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Performance evaluation of AWG-based optical switches with shared wavelength conversion","authors":"A. Pattavina, R. Zanzottera","doi":"10.1109/HSPR.2008.4734455","DOIUrl":"https://doi.org/10.1109/HSPR.2008.4734455","url":null,"abstract":"Since the demand for high-capacity and flexibility in optical WDM networks keeps increasing, the realization of simple optical cross-connects (OXCs) becomes more and more a central topic in the optical research field. In this paper we consider new solutions for the architectures of WDM optical switches employing only arrayed waveguide gratings and tunable wavelength converters. Our interest is focused on single-stage OXC architectures for which non-blocking conditions have been already identified. Here, we go deeper into this subject providing a performance evaluation of such structures in terms of blocking probability, in comparison with other blocking or non-blocking architectures.","PeriodicalId":130484,"journal":{"name":"2008 International Conference on High Performance Switching and Routing","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116094660","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A distributed and scalable MPLS architecture for next generation routers","authors":"K. Nguyen, B. Jaumard","doi":"10.1109/HSPR.2008.4734422","DOIUrl":"https://doi.org/10.1109/HSPR.2008.4734422","url":null,"abstract":"In order to deal with the growing traffic on the Internet while minimizing the router management cost, the next generation routers are gradually replacing the current core routers, which are no longer scalable. They are built with enhanced memory capacity and computing resources, distributed across a very high speed switching fabric. However, the current routing software products, particularly those developed by third-party developers, do not fully exploit the distributed hardware platform of these routers, as they are designed on a centralized software architecture. This paper proposes a distributed software architecture of the MPLS module for the next generation routers. In particular, we investigate the ability to transfer components of the current centralized MPLS architectures on the line cards in order to balance the load between the control card and the line cards. This will improve the robustness, scalability and resiliency of the router. Performance evaluation, in terms of the CPU utilization, is also presented.","PeriodicalId":130484,"journal":{"name":"2008 International Conference on High Performance Switching and Routing","volume":"63 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114096689","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Availability-aware routing in optical networks with primary-backup sharing","authors":"Yonggang Li, Qun Qiu, Longjiang Li","doi":"10.1109/HSPR.2008.4734425","DOIUrl":"https://doi.org/10.1109/HSPR.2008.4734425","url":null,"abstract":"This paper creates a novel availability model for backup resources, based on which we propose a novel availability-aware routing algorithm while the backup resources are allowed to be shared by the incoming requests after computing the availability of the backup resources. For primary-backup sharing, we introduces a comparison of the requested availabilities of every connection being covered by the backup link and the incoming request, based on which we evaluate the availability of the backup link and apply it to computing availability-aware routing. The novelty compared with the previous research cases is that the backup resource availability correlates with the requested availability of incoming connections. Through simulation, our algorithm can get more profitability on availability and better blocking performance.","PeriodicalId":130484,"journal":{"name":"2008 International Conference on High Performance Switching and Routing","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130132587","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"ALS: Aggregated Location Service for routing in MANETs","authors":"K. Khazaei, A. Momeni, S. Mohammadi","doi":"10.1109/HSPR.2008.4734451","DOIUrl":"https://doi.org/10.1109/HSPR.2008.4734451","url":null,"abstract":"One of the biggest challenges in MANETs is the design of scalable routing protocols. With the development and application of position devices, location-based routing has received growing attention. An important component of these protocols is the management of the location information at network nodes, this work is done by location services. In this paper, we present one such scheme called aggregated location service (ALS) which scales well in large mobile ad-hoc networks. Simulation results show that under a specific environment the overhead cost of ALS is asymptotically lower than common methods that had been proposed for MANETs, while its reliability is reasonable.","PeriodicalId":130484,"journal":{"name":"2008 International Conference on High Performance Switching and Routing","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130174752","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Router based detection for low-rate agents of DDoS attack","authors":"D. Nashat, Xiaohong Jiang, S. Horiguchi","doi":"10.1109/HSPR.2008.4734440","DOIUrl":"https://doi.org/10.1109/HSPR.2008.4734440","url":null,"abstract":"The TCP SYN flooding attack is the most prevalent type of DDoS attacks that exhaust network resources. The current detection schemes only work well for the detection of high-rate flooding sources. It is notable, however, that in the current DDoS attacks, the flooding rate is usually distributed among many low-rate flooding agents to make the detection more difficult. Therefore, a more sensitive and fast detection scheme is highly desirable for the efficient detection of these low-rate flooding sources. In this paper, we focus on the low-rate agent and propose a router-based detection scheme for it. The proposed scheme is based on the TCP SYN-SYN/ACK protocol pair with the consideration of packet header information (both sequence and Ack. numbers). To make our scheme more sensitive and generally applicable, the counting bloom filter is used to avoid the effect of SYN/ACK retransmission and the change point detection method is applied to avoid the dependence of detection on sites and access patterns. Extensive trace-driven simulation has been conducted to demonstrate the efficiency of the proposed scheme in terms of its detection probability and also average detection time.","PeriodicalId":130484,"journal":{"name":"2008 International Conference on High Performance Switching and Routing","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134287907","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Constructions and analysis of crosstalk-free optical queues","authors":"Hao-Wei Lan, Cheng-Shang Chang, Jay Cheng, D. Lee","doi":"10.1109/HSPR.2008.4734416","DOIUrl":"https://doi.org/10.1109/HSPR.2008.4734416","url":null,"abstract":"There are several recent advances in the constructions of conflict-free optical queues using optical switches and fiber delay lines (SDL). Many of these conflict-free optical queues are constructed by 2 times 2 switches and thus suffer from the notorious crosstalk problem. To alleviate the crosstalk problem, it is suggested in the literature that one should avoid routing two packets to the same 2 times 2 switch at the same time. A construction with such a property is called crosstalk-free. Our first contribution of this paper is to extend the two-step method by Deng and Lee to a three-step method that converts a conflict-free SDL element into a crosstalk-free SDL element. For certain optical queues, including linear compressors, linear decompressors, and non-overtaking delay lines, an additional step (the fourth step in our paper) can be taken to further reduce the construction complexity. For such optical queues, one only needs one more 2 times 2 switch than the original construction. Our second contribution of this paper is to provide worst case analysis of the signal-to-noise ratio (SNR) in both the original conflict-free SDL element and the converted crosstalk-free SDL element. Closed-form expressions for the worst case SNRs in both SDL elements are obtained. These expressions show that the gain of the worst case SNR in the converted crosstalk-free SDL element is quite substantial.","PeriodicalId":130484,"journal":{"name":"2008 International Conference on High Performance Switching and Routing","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123824993","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}