عنوان مقاله
مقایسه عملکرد TCP با سه پروتکل مسیریابی برای شبکه های تک کاره سیار
فهرست مطالب
مقدمه
کارهای وابسته
پروتکل ها
روش های شبیه سازی
نتایج عملکرد
نتیجه گیری
بخشی از مقاله
تصالات چند گانه TCP
در مورد سورس های چند گانهTCP ، بار ترافیک زمینه 100 Kbps و 200 Kbps از 10 و 40 اتصال CBR را در نظر گرفتیم. گره های فرستنده و گیرنده برای هر اتصال منحصر به فرد بودند، هرچند در برخی موارد، نقطه نهایی TCP نقطه نهایی یک یا چند جریان CBR نیز بود. کارایی های ترکیبی 2، 5 و 10 اتصال TCP با بار ترافیک زمینه 100 Kbps در اشکال 5 و 6 نشان داده شده است.
کلمات کلیدی:
اA Comparison of TCP Performance over Three Routing Protocols for Mobile Ad Hoc Networks Thomas D. Dyer Computer Science Division The Univ. of Texas at San Antonio San Antonio, TX 78249 tdyer@cs.utsa.edu Rajendra V. Boppana Computer Science Division The Univ. of Texas at San Antonio San Antonio, TX 78249 boppana@cs.utsa.edu ABSTRACT We examine the performance of the TCP protocol for bulkdata transfers in mobile ad hoc networks (MANETs). We vary the number of TCP connections and compare the performances of three recently proposed on-demand (AODV and DSR) and adaptive proactive (ADV) routing algorithms. It has been shown in the literature that the congestion control mechanism of TCP reacts adversely to packet losses due to temporarily broken routes in wireless networks. So, we propose a simple heuristic, called fixed RTO, to distinguish between route loss and network congestion and thereby improve the performance of the routing algorithms. Using the ns-2 simulator, we evaluate the performances of the three routing algorithms with the standard TCP Reno protocol and Reno with fixed RTO. Our results indicate that the proactive ADV algorithm performs well under a variety of conditions and that the fixed RTO technique improves the performances of the two on-demand algorithms significantly. 1. INTRODUCTION The TCP protocol has been extensively tuned to give good performance at the transport layer in the traditional wired network environment. However, TCP in its present form is not well-suited for mobile ad hoc networks (MANETs) where packet loss due to broken routes can result in the counterproductive invocation of TCP’s congestion control mechanisms. Although a number of studies have been conducted and protocol modifications suggested, improving TCP performance in MANETs is still an active area of research. The performance studies published to date have either focussed on specific protocols or were restricted to a single TCP connection. The objectives of this study were two-fold: to compare TCP This research has been partially supported by DOD/AFOSR grant F49620-96-1-0472 and NSF grant CDA 9633299. performance over different proposed MANET routing protocols, and to explore the utility of using a sender-based heuristic to distinguish between packet loss due to congestion and loss due to route failures. We conducted simulations to measure the performance of TCP for bulk-data transfers over two of the proposed on-demand protocols, DSR [16] and AODV [18]. A third protocol, ADV [5], was also considered. ADV combines an on-demand approach with proactive distance vector routing. A background UDP traffic load from constant bit rate sources was included so that we could assess the impact of non-TCP traffic on the TCP throughput and vice versa. We examined the relative impact of two existing options designed to enhance TCP performance, selective acknowledgements and delayed acknowledgements. We also considered a modification to the TCP sender designed to lessen the negative effect of TCP’s reaction to retransmit timeouts caused by temporary route failures. Several researchers have designed mechanisms by which the TCP sender is notified of a route failure, and in some schemes, the sender is also notified when the route has been re-established [7, 13]. We have not implemented such a scheme for this study, but our sender-side heuristic gives some indication of the performance gains that can be achieved by altering the TCP sender’s behavior. The results of our simulations demonstrate that the proactive ADV protocol performs well under a variety of conditions. ADV does not, however, benefit from our modifications to the TCP sender. On the other hand, the on-demand algorithms both showed significant performance improvements with the addition of our sender-side heuristic to TCP Reno. 2. RELATED WORK Several performance evaluations of MANET routing protocols for UDP traffic have been presented in the literature [5, 6, 9, 15, 17]. Other studies have presented mechanisms for improving TCP performance in 1-hop wireless networks [3, 4, 2, 20]. Recent studies have addressed the TCP performance problems caused by route failures in an ad hoc network. Chandran et al. proposed a feedback-based scheme called TCPFeedback or TCP-F [7]. In this scheme, when an intermediate node detects the disruption of a route due to the mobility of the next host along that route, it explicitly sends a Route Failure Notification (RFN) to the TCP sender. Upon receiving
