Ioannis Stavrakakis

The LCD interconnection of LRU caches and its analysis

In a multi-level cache such as those used for web caching, a hit at level l leads to the caching of the requested object in all intermediate caches on the reverse path (levels l - 1 ..... 1). This paper shows that a simple modification to this de facto behavior, in which only the l - 1 level cache gets to store a copy, can lead to significant performance gains. The modified caching behavior is called Leave Copy Down (LCD); it has the merit of being able to avoid the amplification of replacement errors and also the unnecessary repetitious caching of the same objects at multiple levels. Simulation results against other cache interconnections show that when LCD is applied under typical web workloads, it reduces the average hit distance. We construct an approximate analytic model for the case of LCD interconnection of LRU caches and use it to gain a better insight as to why the LCD interconnection yields an improved performance.

On the scalability of ad hoc routing protocols

A novel framework is presented for the study of scalability in ad hoc networks. Using this framework, the first asymptotic analysis is provided with respect to network size, mobility, and traffic for each fundamental class of ad hoc routing algorithms. Protocols studied include the following: plain flooding (PF), standard link state (SLS), dynamic source routing (DSR), hierarchical link state (HierLS), zone routing protocol (ZRP), and hazy sighted link state (HSLS). It is shown that PF and ZRP scale better with mobility, SIJS and ZRP scale better with respect to traffic, and HSLS scales better with respect to network size. The analysis provides deeper understanding of the limits and trade-offs inherent in mobile ad hoc network routing. Our analysis is complemented with a simulation experiment comparing HSLS and HierLS. An important contribution of this paper is that HSLS is an scalable, easy-to-implement, alternative to hierarchical approaches for large ad hoc networks.

Distributed Selfish Replication

A commonly employed abstraction for studying the object placement problem for the purpose of Internet content distribution is that of a distributed replication group. In this work, the initial model of the distributed replication group of Leff et al. [CHECK END OF SENTENCE] is extended to the case that individual nodes act selfishly, i.e., cater to the optimization of their individual local utilities. Our main contribution is the derivation of equilibrium object placement strategies that 1) can guarantee improved local utilities for all nodes concurrently as compared to the corresponding local utilities under greedy local object placement, 2) do not suffer from potential mistreatment problems, inherent to centralized strategies that aim at optimizing the social utility, and 3) do not require the existence of complete information at all nodes. We develop a baseline computationally efficient algorithm for obtaining the aforementioned equilibrium strategies and then extend it to improve its performance with respect to fairness. Both algorithms are realizable, in practice, through a distributed protocol that requires only a limited exchange of information.

On the optimization of storage capacity allocation for content distribution

The addition of storage capacity in network nodes for the caching or replication of popular data objects results in reduced end-user delay, reduced network traffic, and improved scalability. The problem of allocating an available storage budget to the nodes of a hierarchical content distribution system is formulated; optimal algorithms, as well as fast/efficient heuristics, are developed for its solution. An innovative aspect of the presented approach is that it combines all relevant subproblems, concerning node locations, node sizes, and object placement, and solves them jointly in a single optimization step. The developed algorithms may be utilized in content distribution networks that employ either replication or caching/replacement. In addition to reducing the average fetch distance for the requested content, they also cater to load balancing and workload constraints on a given node. Strictly hierarchical, as well as hierarchical with peering, request routing models are considered.

On the Effects of Cooperation in DTNs

In a delay tolerant network (DTN) the nodes may behave autonomously deciding on their own whether to implement or not the rules of a routing algorithm. In this paper, the effects of node cooperation (or lack of it) are explored for three well-known routing algorithms proposed for DTNs with respect to the message delivery delay and the transmission overhead incurred until message delivery or the termination of the message spreading process. The results show that the sensitivity of the algorithms to the cooperation degree can be high, to the point of making them inferior to algorithms they typically outperform under a fully cooperative environment. Finally, it is demonstrated how a simple mechanism that incorporates the cooperation degree can help improve effectiveness.

A generic characterization of the overheads imposed by IPsec and associated cryptographic algorithms

This paper presents an assessment of the communication overheads of IPsec and evaluates the feasibility of deploying it on handheld devices for the UMTS architecture. A wide range of different cryptographic algorithms are used in conjunction with IPsec, such as Data Encryption Standard (DES), Advanced Encryption Standard (AES), Message Digest (MD5) and Secure Hash Algorithm 1 (SHA-1). We consider the processing and packetization overheads introduced by these algorithms and quantify their impact in terms of communication quality (added delay for the end-user) and resource consumption (additional bandwidth on the radio interface). We conduct a quantitive analysis based on a detailed simulation model of an IPsec enabled handheld device. We verify our simulation results by comparing against analytic results obtained from an approximate analytic model.

Analysis of a probabilistic topology-unaware TDMA MAC policy for ad hoc networks

The design of an efficient medium access control (MAC) for ad hoc networks is challenging. Topology-unaware time-division multiple-access-based schemes, suitable for ad hoc networks, that guarantee a minimum throughput, have already been proposed. These schemes consider a deterministic policy for the utilization of the assigned scheduling time slots that never utilizes nonassigned slots although in such slots collision-free transmissions are possible even under heavy traffic conditions. A simple probabilistic policy, capable of utilizing the nonassigned slots according to an access probability, fixed for all users in the network, is introduced and analyzed here. The conditions under which the system throughput under the probabilistic policy is higher than that under the deterministic policy are derived analytically. Further analysis of the system throughput is shown to be difficult or impossible for the general case and certain approximations have been considered whose accuracy is also investigated. The approximate analysis determines the value for the access probability that maximizes the system throughput, as well as simplified lower and upper bounds that depend only on a topology density metric. Simulation results demonstrate the comparative advantage of the probabilistic policy over the deterministic policy and show that the approximate analysis successfully determines the range of values for the access probability for which the system throughput under the probabilistic policy is not only higher than that under the deterministic policy, but it is also close to the maximum.

Quality of Future Internet Services

This book is the final report of COST – Cooperation europeenne dans le domaine de la recherche scientifique et technique , Action 263 – Quality of future Internet Services.

A comparative evaluation of intrusion detection architectures for mobile ad hoc networks

Mobile Ad Hoc Networks (MANETs) are susceptible to a variety of attacks that threaten their operation and the provided services. Intrusion Detection Systems (IDSs) may act as defensive mechanisms, since they monitor network activities in order to detect malicious actions performed by intruders, and then initiate the appropriate countermeasures. IDS for MANETs have attracted much attention recently and thus, there are many publications that propose new IDS solutions or improvements to the existing. This paper evaluates and compares the most prominent IDS architectures for MANETs. IDS architectures are defined as the operational structures of IDSs. For each IDS, the architecture and the related functionality are briefly presented and analyzed focusing on both the operational strengths and weaknesses. Moreover, methods/techniques that have been proposed to improve the performance and the provided security services of those are evaluated and their shortcomings or weaknesses are presented. A comparison of the studied IDS architectures is carried out using a set of critical evaluation metrics, which derive from: (i) the deployment, architectural, and operational characteristics of MANETs; (ii) the special requirements of intrusion detection in MANETs; and (iii) the carried analysis that reveals the most important strengths and weaknesses of the existing IDS architectures. The evaluation metrics of IDSs are divided into two groups: the first one is related to performance and the second to security. Finally, based on the carried evaluation and comparison a set of design features and principles are presented, which have to be addressed and satisfied in future research of designing and implementing IDSs for MANETs.

Adaptive playout strategies for packet video receivers with finite buffer capacity

Due to random delay variations in current best effort networks, packet video applications rely on end-system buffering and playout adaptation to reduce the effects of disruptions on the required smooth stream presentation. To study the effect of buffering and playout adaptation, we present an analytical model based on the M/G/1 queueing system with finite buffer capacity, and traffic intensity equal to or greater than unity. This model fits well a range of new applications that have limited buffer resources for the reception of incoming frames. We introduce the variance of distortion of playout (VDoP), a new metric that accounts for the overall presentation disruption caused by buffer underflows, intentionally introduced gaps during slowdown periods and data loss from overflows. VDoP is an elegant and fair metric for the estimation of playout quality and will hopefully assist the development of better adaptation algorithms. Furthermore, the effect of finite buffer capacity is examined in relation to stream continuity revealing a system behavior not previously accounted for. The sensitivity of the system to the variance of the arrival process is also examined by means of simulation. Finally, an online algorithm is presented for the exploitation of our study on implemented systems.