Catálogo de publicaciones - libros

The TCP performance degradation over heterogeneous networks is caused by not only network congestion, but also random errors of wireless links. We propose an end-to-end stable accurate rapid bandwidth estimate (SARBE) algorithm reacting appropriately to the end-to-end loss differentiation estimate algorithm (LDED), which can improve the TCP performance over heterogeneous networks without the incipient congestion notifications of the intermediate routers. LDED detects the signal of incipient congestion to lead the sender to enter the congestion avoidance phase opportunely before router’s queue overflows. As well as relying distinguishing ability the causes of loss, our algorithm adjusts the packet transmission rate precisely according to the estimated bandwidth after new ACK receiving, fast retransmit or transmission timeout events.

In many fields of engineering and science, researchers and engineers use computers to simulate natural phenomena rather than conducting experiments involving the real system. With the power of today’s computers, simulation provides an easy way of predicting the performance of a prospective system or comparing several alternatives at the system design stage. In this paper, we present a simulation framework specifically designed for wireless Internet access networks. This simulation framework is designed for the three protocol layers, HTTP, TCP/IP and the link-layer protocol, and can be configured independently in each of these layers. The time-driven nature of the framework allows us to observe the diurnal changes of the system in the simulation, which in turn makes it possible to evaluate the statistical properties of the system.

As a new technique, one characteristic of wireless sensor networks (WSNs) is their limited system lifetime. Therefore, it is more important to save energy and proportion energy consumption. This paper presents a weight-directed based multi-hop routing algorithm for WSNs. This algorithm can transfer data quickly to goal sensor node using the directional information and RWVs (route weight value) of sensor nodes as well as balance energy consumption of all sensor nodes. Detailed simulations of sensor network environments indicate that this algorithm improves energy efficiency and proportions energy consumption of all sensor nodes to extend network system lifetime, and routes data quickly in comparison to the flooding algorithm.

Quality of service (QoS) and security in Mobile IP networks are becoming significant issues due to an increasing number of wireless devices[1]. For this reason, the Hierarchical Mobile IPv6 (HMIPv6) protocol[2] and the Authentication, Authorization, and Accounting (AAA) protocol[3] are proposed. However this protocol has inefficient authenticating and binding update procedures that limit its QoS. In this paper, we propose a forwarding scheme extension for fast and secure handoff that can reduce a handoff delay while maintaining a security level by a forwarding and session key exchange mechanism. The performance results show that the proposed mechanism reduces the handoff latency up to 10% and the handoff failure rate up to 25% compared to the previous mechanism.

We propose a distributed and adaptive topology management algorithm, called PATM (Priority-based Adaptive Topology Management), that constructs and maintains a connected backbone topology based on a minimal dominating set of the network. PATM provides a succinct presentation of the network topology to routing protocols, and therefore reduces the control overhead in routing updates. Two optimizations are proposed to further reduce the topological information exchanges among the nodes by piggybacking topology updates in packets transmitted by each node, and adaptively adjusting the topology update intervals. The efficiency of the algorithm is validated by simulations based on DSR (Dynamic Source Routing) protocol. The simulation results demonstrate that PATM not only significantly reduces the routing control overhead, but also substantially improves the network data forwarding performance.

Intrusion Detection Systems (IDSs) for Mobile Ad hoc NETworks (MANETs) is becoming an exciting and important technology in very recent years, because the intrusion prevention techniques can not satisfy the security requirements in mission critical systems. The proposed IDS architecture can be divided into two categories by the distributing form of IDS agents: fully distributed IDS and cluster-based IDS. The former has a high detection ratio, but it also consumes a cascade of energy. The latter has considered energy saving, but some hidden troubles of security exist in it. In this paper, we have proposed a novel IDS Agent Distributing (IAD) protocol for distributing IDS agents in MANETs. IAD protocol divides the whole network into several zones, selects a node subset from each zone, and runs IDS agent on the node in this subset. At the same time, IAD protocol can rectify the number of nodes running IDS agent according to the threat level of the network. Compared with the scheme that each node runs its own IDS, our proposed scheme is more energy efficient while maintaining the same level of detection rate. While compared with the cluster-based IDS scheme, our scheme is more flexible when facing the emergent situations. Simulation results show that our scheme can effectively balance the security strength and energy consuming in practice.

Due to an increasing number of portable devices, a support for quality of service (QoS) and security becomes an main issue in Mobile IP networks. However Authentication, Authorization, and Accounting (AAA) protocol has inefficient authenticating procedures that limit its QoS. That is, a mobile node (MN) should be distributed new session keys whenever it performs a handoff. As computing power of mobile devices becomes greater, a key distribution using a symmetric key cannot guarantee the security. Hence, we employ an ID-based cryptography to intensify the security and when the MN moves to a new domain, a foreign agent (FA) reuses previous session keys encrypted by a public key for the fast handoff. Our proposed scheme reduces handoff delay and maintains high security by exchanging previous session keys between FAs. The performance results show that the proposed scheme reduces the latency up to about 63% compared to the previous ID-based AAA.

This paper discussed the resource allocation and reservation for wireless network, which is a challenging task due to the mobility uncertainty of user, Motivated from a rationale, i.e., a good data compressor should be a good predictor, we proposed a mobility prediction algorithm. Integrating the prediction algorithm into GC, a resource allocation scheme is also proposed. The numerical simulation results show that the time-complexity of our proposed scheme is worse, but it outperforms and in the QoS support effectiveness.

Handover is one of the most important factors that may degrade the performance of TCP connections and real-time applications in wireless data networks. We developed a loss-free handover scheme called LPM (Last Packet Marking) for IEEE 802.16e-based broadband wireless access networks. By integrating MAC and network layer handovers efficiently, LPM minimizes the handover delay and eliminates packet losses during handover. Our performance study shows that LPM achieves loss-free packet delivery without packet duplication and increases TCP throughput significantly.

We study the coverage problem from the fault tolerance point of view for sensor networks. Fault tolerance is a critical issue for sensors deployed in places where are not easily replaceable, repairable and rechargeable. The failure of one node should not incapacitate the entire network.

We propose three 1 fault tolerant models, and we compare them among themselves, and with the minimal coverage model [8].