Search

The concept of protocol is an intellectual terrain on which one might contemplate a number of overlapping, sometimes contradictory and often interrelated theoretical problems at play in modern times. A study describes a few historical waypoints for the concept, identifies some of the normative claims inherent in using the term, and also suggests some issues with which, going forward, any theorist of computers or networks has to contend.

Hybrid Fiber Coaxial (HFC) networks provide a wide bandwidth and represent the solution of choice for many residential networks nowadays. The Data Over Cable Service Interface Specification (DOCSIS) protocol is an important standard for HFC networks and is supported by the majority of current vendors. This protocol uses the Truncated Binary Exponential Back-off (TBEB) algorithm to resolve collisions within the network by means of a back-off window. However, the performance of this algorithm tends to deteriorate when the network load is high. Consequently, the present study develops a novel mathematical model for the TBEB algorithm and then uses this model to identify the window size which yields the optimal system throughput and minimum delay time under high traffic conditions. The present numerical results confirm that the performance is improved when the window settings identified from the developed model are applied. Index Terms--CATV, DOCSIS protocol, Hybrid Fiber Coaxial (HFC), MAC.

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.comnet.2005.05.018 Byline: Nizar Bouabdallah (a), Andre-Luc Beylot (b), Guy Pujolle (a) Keywords: Optical access network; Medium Access Control (MAC) Protocol; Fairness control; Access delay evaluation Abstract: The packet-based optical access ring is becoming a promising solution in metropolitan networks. Its performance depends mainly on how optical resource sharing takes place among the different competing access nodes. In such a network fairness issues are likely to arise between upstream and downstream nodes competing to access a common data channel. This paper provides a new protocol called TCARD (Traffic Control Architecture using Remote Descriptors), based on a preventive reservation scheme, to alleviate the fairness problems. The reservation process consists in forcing the upstream nodes to ensure enough available bandwidth for downstream nodes' transmission purpose. Moreover, we develop analytical models in order to assess the access delay when TCARD is enabled. The analytical and simulation results show how the proposed solution alleviates the performance degradation and the resource sub-utilization while achieving fairness among bus nodes. Author Affiliation: (a) LIP6, University of Paris 6, 8 rue du Capitaine Scott, F-75015 Paris, France (b) ENSEEIHT, IRIT/TeSA Lab. 2, rue C. Camichel, BP7122, F-31071 Toulouse, France

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.comnet.2005.02.009 Byline: N. Bartolini (a), E. Casalicchio (b) Keywords: Context transfer protocol; IPv6; Fast handover; Performance analysis Abstract: In nowadays wireless networks, mobile users frequently access Internet services that are often based on information concerning the application context and service status. In presence of mobility, the procedure of service handover, may require a restart of the ongoing service, if the necessary context information is not properly transferred to the new point of access. Context transfer procedures introduce additional overheads to handovers possibly affecting the quality of service perceived by mobile users and making handovers very critical. In this paper the need for efficient protocols for transferring service context and profile related information is pointed out with reference to many mobile internet services, and the possible scenarios are differentiated on the basis of the handover triggering mechanisms. A performance model to compare these mechanisms, when context transfer protocols run on top of IPv6 with fast handover, is proposed. Numerical results point out the necessity to adapt the triggering mechanism to the size of the context data. Author Affiliation: (a) Department of Computer Science, University of Rome "La Sapienza", Via Salaria 113, Rome 00198, Italy (b) Department of Computer Science, University of Rome "Tor Vergata", Rome 00133, Italy Article History: Received 10 January 2005; Revised 10 January 2005; Accepted 17 February 2005 Article Note: (miscellaneous) Responsible Editor: Dr. G. Morabito

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.compeleceng.2006.01.016 Byline: Ing-Chau Chang (a), Yi-Ting Mai (b) Keywords: IP Multicast; Mobile IP; Multimedia QoS; Heterogeneous wireless networks; SIP; HMSIP Abstract: For supporting the mobile host (MH) to be the multicast receivers on wireless networks, the traditional approaches such as the Hierarchical Remote Subscription (HRS) and Bi-directional Tunneling with Mobile Multicast (BT-MoM) integrate the Mobile IP protocol with the IP multicast technique. However, they inherit intrinsic problems to handle mobility and suffer significant delays to rebuild the multicast tree when the MH moves out of the multicast tree, which results in serious playback interruption and quality of service (QoS) degradation for the ongoing multimedia service especially when the MH are scattered over different kinds of wireless networks. In this paper, we propose a cross-layer framework and the Hierarchical Multicast Session Initiation Protocol (HMSIP), which extend the IETF Session Initiation Protocol (SIP) with the concept of multicast SIP session and integrate the underlying IP Multicast QoS routing protocol and the RSVP resource reservation technique, to support mobile multimedia multicasts. This framework significantly reduces handoff delays and deployment costs and hence achieves end-to-end QoS support for MHs on heterogeneous wireless networks. Further, we mathematically analyze handoff delays of the HRS, BT-MOM and HMSIP schemes. Moreover, simulation results show that the HMSIP can significantly achieve much more stable and lower handoff delays than those of the BT-MOM and HRS only with a small proportion of extra bandwidth than the bandwidth-optimal HRS, thereby supporting higher numbers of handoffs without interrupting multimedia playback of the mobile multicast activity. Tradeoffs between costs for deploying HMSIP Proxies in the framework and corresponding performances of the HMSIP families are discussed. Finally, multimedia QoS improvements of MHs are shown to emphasize advantages by integrating RSVP in this HMSIP framework. Author Affiliation: (a) Department of Computer Science and Information Engineering, National Changhua University of Education, Changhua, Taiwan, ROC (b) Department of Computer Science and Information Engineering, National Chi Nan University, Taiwan, ROC

In this paper, the session setup delay of the session initiation protocol (SIP) is studied. The transmissions on both the forward and reverse channel are assumed to experience Markovian errors. The session setup delay is evaluated for different transport protocols, and with the use of the radio link protocol (RLP). An adaptive retransmission timer is used to optimize SIP performances. Using numerical results, we find that SIP over user datagram protocol (UDP) instead of transport control protocol (TCP) can make the session setup up to 30% shorter. Also, RLP drastically reduces the session setup delay down to 4 to 5 s, even in environments with high frame error rates (10%) and significant correlation in the fading process ([f.sub.D] T = 0.02). SIP is compared with its competitor H.323. SIP session setup delay with compressed messages outperforms H.323 session setup delay. Index Terms--Correlated fading channels, IP-based wireless networks, radio link protocol (RLP), session initiation protocol (SIP), session setup delay, transport control protocol (TCP), user datagram protocol (UDP).

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.comnet.2005.05.015 Byline: Tianwei Chen (a), Michel Sortais (b), Gunter Schafer (a), Stefan Adams (c), Changpeng Fan (d), Adam Wolisz (a) Keywords: QoS; Mobility; DoS; Cookie protection; Performance analysis Abstract: Quality of Service (QoS) mechanisms in networks supporting mobile Internet communications give rise to Denial of Service (DoS) threats: if the network cannot efficiently check the credibility of a QoS request during a handover process, malicious entities could flood the network with bogus QoS requests; if the authentication check is performed by means of an AAA protocol before the access network commits its resources, the authentication process may not only introduce a notable latency to the handover process, but also generate an extensive traffic in the presence of malicious requests, thus causing the network signaling capacity to degrade. In order to defend against these kinds of attacks and meet the low-latency micro-mobility handover requirement, we propose a preliminary authentication check with a cookie-based mechanism before processing the requests and performing authentication and authorization. Our performance evaluation shows that the cookie-based mechanism is efficient in dealing with the identified issues. Author Affiliation: (a) Fachgebiet Telekommunikationsnetze, Technische Universitat Berlin, Germany (b) Math. Institut, Fak. II MA 7-4, Technische Universitat Berlin, Germany (c) Max Planck Institute for Mathematics in the Sciences, Leipzig, Germany (d) ICM N PG SP RC PN, Siemens AG, Germany

In this paper, we consider distributed systems that can be modeled as finite state machines with known behavior under fault-free conditions, and we study the detection of a general class of faults that manifest themselves as permanent changes in the next-state transition functionality of the system. This scenario could arise in a variety of situations encountered in communication networks, including faults occurred due to design or implementation errors during the execution of communication protocols. In our approach, fault diagnosis is performed by an external observer/diagnoser that functions as a finite state machine and which has access to the input sequence applied to the system but has only limited access to the system state or output. In particular, we assume that the observer/diagnoser is only able to obtain partial information regarding the state of the given system at intermittent time intervals that are determined by certain synchronizing conditions between the system and the observer/diagnoser. By adopting a probabilistic framework, we analyze ways to optimally choose these synchronizing conditions and develop adaptive strategies that achieve a low probability of aliasing, i.e., a low probability that the external observer/diagnoser incorrectly declares the system as fault-free. An application of these ideas in the context of protocol testing/classification is provided as an example. Index Terms--Communication protocols, discrete event system (DES), distributed systems, finite-state machines (FSMs), Markov models, probabilistic fault detection.

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.sysarc.2005.01.001 Byline: L. Lo Bello, A. Gangemi Keywords: Real-time communication; Slotted rings; Spatial reuse; Earliest Deadline First; Swapping policies Abstract: This paper proposes an architecture and communication protocol called the Slot Swapping Protocol which is used to interconnect, by means of a backbone, different subsystems that generate and consume real-time data. The proposed communication architecture fits well into the general Computer Integrated Manufacturing (CIM) model, thus representing an appealing solution for FieldBus interconnection in factory automation systems. Inspired by the concept of session swapping, the Slot Swapping Protocol makes a step forward in slotted rings by including explicit time constraints in traffic scheduling, as it assigns a dynamic priority to slots and uses an Earliest Deadline First-based swapping policy. The paper describes the communication protocol, discusses three slot swapping approaches and presents a set of experiments characterizing the protocol's ability to meet traffic deadlines in several working conditions. Author Affiliation: Department of Computer Engineering and Telecommunications, University of Catania, Italy Article History: Received 10 March 2003; Revised 24 February 2004; Accepted 11 January 2005

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.sysarc.2002.10.001 Byline: S.W. Ng, Edward Chan Keywords: Real-time streaming protocols; Congestion control; Wireless networks; TCP friendliness Abstract: A major problem in supporting multimedia streaming in the Internet is that the streaming protocol used tends to take bandwidth away from competing TCP traffic streams. Thus it is important to devise rate based protocols that are TCP-friendly. In the paper we examine the effectiveness of two important models in devising TCP-friendly rate based protocols in lossy environment such as wireless networks. Using simulation for a variety of network scenarios, it is shown that the rate based protocols based on one model underestimates while the other overestimates TCP bandwidth in the presence of errors. A modification is made to the models which allows us to construct a rate based protocol that can track TCP bandwidth better in the presence of random error and hence exhibit behavior that is more TCP-friendly. Author Affiliation: Department of Computer Science, City University of Hong Kong, 83 Tat Chee Ave., Kowloon, Hong Kong Article History: Received 8 January 2002; Revised 4 July 2002; Accepted 15 October 2002

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.comnet.2004.11.020 Byline: Jun He, Hung Keng Pung Keywords: Fairness; Max-min fairness index; Multi-hop; Ad hoc wireless networks Abstract: It is an undisputed fact that fairness is an important element of a well-designed medium access control (MAC) protocol for multi-hop ad hoc networks. However, most popular MAC protocols still fail to attain an acceptable level of fairness in media access although several enhancements have been proposed in the past. These proposed enhancements are effective only in limited scenarios. It is our objective in this paper to do the following: (i) analyze the fairness problem; (ii) identify and analyze the three main causes leading to the fairness problem, namely, the lack of synchronization problem (LSP), the double contention areas problem (DCP) and the lack of coordination problem (LCP); (iii) based on the analysis, propose a new MAC protocol named the extended hybrid asynchronous time division multiple access (EHATDMA) as a solution. For better assessment of fairness, we have designed an index named max-min fairness index, which is scenario-independent and reflects the difference between the fair sharing provided by a protocol and the ideal max-min fair sharing. Comprehensive simulations have been carried out to compare the fairness of our protocol with the existing ones. Simulation results show that although the existing protocols employ various enhancements meant to improve the fairness property, most of them are still strongly biased towards throughput when a conflict between throughput and fairness arises. In addition, the fairness performance of these protocols varies widely from one scenario to another. On the other hand, EHATDMA strikes a good balance between throughput and fairness. It delivers a consistently high level of fairness regardless of network topology, traffic load and radio parameters, yet maintains high throughput whenever possible. Our simulation results also reveal that the most important mechanism affecting the fair sharing of radio channels among flows is the non-work-conserving mechanism. Author Affiliation: School of Computing, National University of Singapore, 3 Science Drive 2, Singapore S117543, Singapore Article History: Received 20 February 2004; Revised 10 August 2004; Accepted 5 November 2004 Article Note: (miscellaneous) Responsible Editor: L. Lenzini

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.compeleceng.2005.03.001 Byline: Her-Tyan Yeh (a), Hung-Min Sun (b) Keywords: Network security; Authentication; Key exchange; Password guessing attacks; Perfect forward secrecy; Cryptography Abstract: Up to now, all papers in password-authenticated key exchange protocols are constrained to two-party or three-party models. Under these two models, the clients are registered or authenticated by the same server. However, in reality two communicants may register under two different servers. In this case, the above models are inefficient or not suitable any more. This paper will discuss password authenticated key exchange protocol where both communicants are registered in two distinct servers. Author Affiliation: (a) Department of Information and Communication, Southern Taiwan University of Technology, Tainan 710, Taiwan (b) Department of Computer Science, National Tsing Hua University, Hsinchu, Taiwan 300, Taiwan Article History: Received 15 September 2004; Revised 1 February 2005; Accepted 9 March 2005

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.comnet.2004.08.002 Byline: Chunmei Liu, Eytan Modiano Keywords: AIMD protocol; TCP; ARQ; Hybrid networks; Interactions Abstract: It is well known that TCP does not perform well in networks with large propagation delay and high link loss probability, which are typical in hybrid space-terrestrial networks. Implementing Automatic-Repeat-Request (ARQ) retransmission over satellite links in hybrid networks has been proposed to enhance TCP performance by hiding the link losses from TCP. This paper analyzes the impact of implementing ARQ over satellite links on TCP performance. In particular, this paper focuses on the window flow control mechanism of TCP, and provides an exact model for a hybrid space-terrestrial system with transport layer Additive-Increase-Multiplicative-Decrease (AIMD) protocols and satellite link layer ARQ. Both Go-Back-N (GBN) and Selective-Repeat (SRP) are considered, and the delay of acknowledgements (ACKs) of ARQ are taken into account. The numerical solutions for the throughput as a function of different protocol and packet loss parameters are presented. Both the analysis and numerical results show that in most cases, implementing ARQ over satellite links can significantly improve TCP performance, and that proper choices of the protocol parameters can significantly improve the system performance as well. Although the paper focuses on hybrid space-terrestrial networks, the analysis and results are also applicable to other networks with large bandwidth-delay product and one error-prone bottleneck link. Author Affiliation: Massachusetts Institute of Technology, Cambridge, MA 02139, United States Article Note: (footnote) [star] This work was supported by NASA Space Communication Project grant number NAG3-2835.

A novel on-demand protocol called Backup Source Routing (BSR) is proposed to establish and maintain backup routes that could be utilized after the primary path breaks. The key advantage of BSR is the reduction of the frequency of route discovery flooding, which is recognized as a major overhead in on-demand protocols.

Byline: Lichun Bao (001), J. J. Garcia-Luna-Aceves (001) Keywords: ad hoc networks; directional antenna; channel access scheduling; multi-beam adaptive array (MBAA) Abstract: Directional antennas can adaptively select radio signals of interest in specific directions, while filtering out unwanted interference from other directions. A couple of medium access protocols based on random access schemes have been proposed for networks with directional antennas, using the omnidirectional mode for the transmission or reception of control packets in order to establish directional links. We propose a distributed receiver-oriented multiple access (ROMA) scheduling protocol, capable of utilizing multi-beam forming directional antennas in ad hoc networks. Unlike random access schemes that use on-demand handshakes or signal scanning to resolve communication targets, ROMA computes a link activation schedule in each time slot using two-hop topology information. It is shown that significant improvements on network throughput and delay can be achieved by exploiting the multi-beam forming capability of directional antennas in both transmission and reception. The performance of ROMA is studied by simulation, and compared with a well-know static scheduling scheme that is based on global topology information. Author Affiliation: (001) Computer Engineering Department, School of Engineering, BE 399, University of California, Santa Cruz, CA, 95064, USA Article History: Registration Date: 13/12/2004

Byline: Eun-Sun Jung (001), Nitin H. Vaidya (002) Keywords: power control; energy saving; medium access control; ad hoc network Abstract: This paper presents a power control MAC protocol that allows nodes to vary transmit power level on a per-packet basis. Several researchers have proposed simple modifications of IEEE 802.11 to incorporate power control. The main idea of these power control schemes is to use different power levels for RTS--CTS and DATA--ACK. Specifically, maximum transmit power is used for RTS--CTS, and the minimum required transmit power is used for DATA--ACK transmissions in order to save energy. However, we show that these schemes can degrade network throughput and can result in higher energy consumption than when using IEEE 802.11 without power control. We propose a power control protocol which does not degrade throughput and yields energy saving. Author Affiliation: (001) Department of Computer Science, Texas A&M University, College Station, TX, USA (002) Department of Electrical and Computer Engineering, and Coordinated Science Laboratory, University of Illinois, Urbana, USA Article History: Registration Date: 13/12/2004

In future wireless ad hoc networks, transactions are expected to become one of the primary types of flows. Transactions require only a small number of packets to complete. Hence, providing optimal (shortest path) routes to such transactions consumes more energy than the actual data transfer. Conventional shortest path routing protocols are, thus, unsuitable for routing transactions. In this paper, we present a novel architecture, called TRANSFER, for transactions routing in large-scale wireless ad hoc networks. In our approach, we aim to reduce the total energy consumption by transactions as opposed to finding shortest path routes. Our architecture uses a hybrid approach, in which each mobile node obtains information about nodes in its proximity (zone), up to R hops away, using a proactive link-state protocol. Beyond the proximity, we introduce the novel notion of contacts that act as shortcuts to reduce the degrees of separation between the source of the transaction and the destination. We propose an efficient on-demand protocol for contact selection that does not assume knowledge of location information. Contacts are used during transactions and queries to discover valid routes in an energy-efficient manner. Extensive simulations are used to evaluate the performance of our protocol in terms of energy consumption and success rate. We compare our architecture to flooding, dynamic source routing (DSR), zone-routing protocol (ZRP), and two power-aware schemes. Our results show substantial power savings for our contact-based protocol, especially for large ad hoc networks. Index Terms--Adhoc networks, contact-based architecture, energy efficiency, network simulation, routing protocol.

A novel hybrid routing protocol - the Two-Zone Routing Protocol (TZRP) as a non-trivial extension of Zone Routing Protocol is proposed. TZRP uses two different zones to decouple the protocols ability to adapt to traffic characteristics from the ability to adapt to mobility.

Two of the fundamental issues in designing protocols for message passing between mobile agents (MAs) are tracking the migration of the target agent and forwarding messages to it. Even with an ideal fault-free network-transport mechanism, messages can be dropped during MA migration. Therefore, in order to provide reliable message delivery, protocols need to overcome message loss caused by asynchronous operations of agent migration and message forwarding. In this paper, two known message forwarding approaches, namely push and pull, are explored to design adaptive and reliable message delivery protocols. Based on a commonly used MA tracking model, the pros and cons of these two approaches are evaluated, both qualitatively and quantitatively. The comparative performance evaluation is presented in terms of network traffic and delay in message processing. We also propose improvements to the pull approach to reduce network traffic and the message delay. We conclude that with different message passing and migration patterns and varying requirements of real-time message processing, specific applications can select different message delivery approaches to achieve the desired level of performance and flexibility. Index Terms--Mobile agents (MAs), pull, push, reliable message delivery.

A novel, efficient and low cost start-up and restart synchronization approach for TDMA environments is presented by focusing on bus-based communication protocols. The approach utilizes information about a node's message length that forms a unique sequence to achieve synchronization such that communication overhead can be avoided.

Different static and dynamic schemes are explored to select the virtual lanes for outgoing messages and mix the load among them without restricting any particular type of message to be carried only by a particular virtual network. In addition, the network latency, the number of message buffers per virtual lane, and the depth of network interface output queues affect the relative performance of various virtual lane selection algorithms are examined.

Byline: Yuan Xue (1), Klara Nahrstedt (1) Keywords: ad hoc networks; routing protocol; misbehaving node Abstract: Most existing designs of ad hoc networks are based on the assumption of non-adversarial environments, where each node in the network is cooperative and well-behaved. When misbehaving nodes exist in the network, the performance of current routing protocols degrades significantly. Since ad hoc networks, consisting of autonomous nodes, are open and distributed in nature, maintaining a fault-free network environment is extremely difficult and expensive. In this paper, we propose a new routing service named best-effort fault-tolerant routing (BFTR). The design goal of BFTR is to provide packet routing service with high delivery ratio and low overhead in presence of misbehaving nodes. Instead of judging whether a path is good or bad, i.e., whether it contains any misbehaving node, BFTR evaluates the routing feasibility of a path by its end-to-end performance (e.g. packet delivery ratio and delay). By continuously observing the routing performance, BFTR dynamically routes packets via the most feasible path. BFTR provides an efficient and uniform solution for a broad range of node misbehaviors with very few security assumptions. The BFTR algorithm is evaluated through both analysis and extensive simulations. The results show that BFTR greatly improves the ad hoc routing performance in the presence of misbehaving nodes. Author Affiliation: (1) Department of Computer Science, University of Illinois at Urbana-Champaign, Urbana, Illinois, 61801, U.S.A. Article History: Registration Date: 29/12/2004