Catálogo de publicaciones - libros

Multi-hop wireless networks provide a quick and easy way for networking when we need a temporary network or when cabling is difficult. The 802.11 Medium Access Control (MAC) plays an important role in the achievable system performance. There have been many studies on analytic modeling of single-hop 802.11 wireless networks but only a few on the analysis of multi-hop wireless networks. Furthermore, the object of these researches is an homogeneous ad-hoc wireless networks; therefore they are not appropriate for a network with structure such as wireless mesh networks. This paper introduces an analytic model of throughput performance for the IEEE 802.11 multi-hop networks, which allows us to compute the achievable throughput on a given path in multi-hop wireless networks. The model shows that there is an optimal point at which throughput is maximized. Using this model and a Markov model for modeling the operation of the IEEE 802.11 DCF we can determine the amount of data that each node should inject to the network to get the best throughput performance.

In large-scale wireless sensor networks, efficient service discovery and data transmission mechanisms are both essential and challenging. Ant colony algorithm which has been used to resolve routing, localization and object tracing issues in mobile ad hoc and sensor networks provide a valuable solution for this problem. In this paper, we describe a novel scalable Action-based Service Discovery Protocol (ASDP) using ant colony algorithm in wireless sensor networks. ADSP can abstract the semantics information from the data via the nodes or user operation and map them into six different action sets. Then it adjusts the related parameters to satisfy with service and transmission requirements from different kinds of actions. We evaluate it against other approaches to identify its merits and limitations. The simulation results show that ASDP can maximize the network utilization. Farther experiments indicate it scales to large number of nodes.

In this paper, we propose a real time scheduling policy over 802.11 DCF protocol called Deadline Monotonic (DM). We evaluate the performance of this policy for a simple scenario where two stations with different deadlines contend for the channel. For this scenario a Markov chain based analytical model is proposed. From the mathematical model, we derive expressions of the saturation throughout and the average medium access delay called service time. Analytical results are validated by simulation results using the ns-2 network simulator.

With the network size growing, routing protocols of mobile ad-hoc networks (MANET) face many challenges. The most critical issue is the lack of bandwidth and computation capability. This paper describes the Parallel Link State Routing protocol (PLSR) that is suitable for MANET. The key concept used in the protocol is that of graph decomposition. This technique substantially raises the rate of routing updating; meanwhile it reduces the message overload as compared with a classical flooding mechanism and some other link state routing protocols. In PLSR, the network is a two-levels topology, and many problems caused by “flat” multi-hoppings will disappear. Thus, the second optimization is achieved by sub-network that reduces the control messages. As a third optimization, the routing computation will be in several sub-networks simultaneity. Hence, comparing with current link state algorithms, the routing table of PLSR will be regenerated quickly as the network topology changes.

It has been shown analytically [1],[2] that significant performance improvements as compared to existing technologies (e.g., IEEE 802.11) can be achieved in random access wireless networks. In [3] we proposed a fully distributed channel access paradigm based on the opportunistic communication principal called the Channel MAC paradigm suitable for distributed wireless networks such as ad hoc networks. In this paper, we analytically derive the throughput of the Channel MAC. It provides a throughput-limit on the channel-based MAC mechanism in shared multiple access environments without collisions or capturing effects. Both simulation and analytical results reveal possible performance improvement over existing techniques.

Traditional transmission control protocol reduces its performance by misinterpreting mobility losses due to node motion as congestion losses in wireless mobility network. While it has been shown that TCP Vegas provides better performance compared to TCP Reno, studies have identified various issues associated with the protocol in mobile ad hoc networks. This paper proposed adaptive congestion control scheme based on Traditional TCP. The key idea of our approach is to monitor permanently the TCP flow and record useful data to infer the current state of mobility ad hoc networks when packet losses are perceived, distinguish losses packets between bit error and congestion based on fuzzy logic theory, then based on wireless network link state, such as bit error, network congestion, or rerouting, Appropriate congestion control algorithm be choose to overcome limitations. Our experiments show that the scheme is able to avoid oscillation of TCP sender congestion window, and obtain more throughput than traditional transmission control protocol.

Recently, some proposals have suggested that maintaining the same contention window (CW), or reducing it, for nodes suffering packet losses, due to channel transmission impairments, is effective in increasing the performance of the IEEE 802.11 in noisy environment. Our study presented inhere will prove analytically and via simulations that this should not be necessarily the case. To facilitate our analysis, we consider two binary exponential backoff (BEB) algorithms in our study: a standard BEB where a host increases its CW upon every packet loss (collision or transmission error) and another access method with a capability to differentiate between the type of losses; here, a host experiencing a loss will increase its CW only after a collision and remain in the same backoff stage otherwise. We show that the second access procedure outperforms the standard BEB when the network is lightly loaded. However, in a congested network, this quick recovery property results in intensifying the collisions among contending nodes and hence yields a poor system performance. We propose a hybrid method that takes advantage of both access methods to achieve better throughput under various network loads.

The choice of transmit power determines the transmission range and carrier sensing range. Thus, the transmit power is a key to the tradeoff between the amount of spatial reuse and probability of collisions in wireless multi-hop ad hoc networks. In this paper, we propose a realistic analytical model to study the performance of the IEEE 802.11 access method in multi-hop wireless networks by incorporating the effect of transmit power and physical carrier sense. Using the model, we investigate the effect of these attributes on the throughput obtained by individual nodes. Our results show that performing power control and carrier sense threshold tuning using the two way handshake offer higher advantages and achieve better throughput than that achieved using the four-way handshake.

The access scheduling on the control channels in TDMA wireless mesh networks is studied in this paper. The problem is to assign time-slots for each node in the network to access the control channels so that it is guaranteed that each node can transmit a control packet to any one-hop neighbor in the scheduling cycle. The objective is to minimize the total number of different time-slots in the cycle. This paper has taken the large interference range problem into consideration and proposed two algorithms for the scheduling problem, namely, One Neighbor Per Cycle (ONPC) algorithm and All Neighbors Per Cycle (ANPC) algorithm. The number of time-slots by ANPC algorithm is upper-bounded by min(, 42) in some special cases, where is the node numberand is the maximum node degree. Both centralized and fully distributed versions are given. Simulation results also show that the performance of ONPC is rather better than ANPC.

This paper considers large-scale dense wireless sensor networks with path-constrained mobile sink. From the viewpoints of improving data equilibrium and energy-saving, we propose an enhanced energy efficient communication protocol in which a partition algorithm of overlapping time based on data equilibrium is designed to minimize the data variance and disequilibrium from each node. We present a data equilibrium information diffusion method to control the amount of collected data to avoid the waste of data and energy. Simulations under OMNET++ and MATLAB show that with protocol and algorithms proposed in this paper we can achieve better performance than other conventional methods in terms of data equilibrium rate, network lifetime and energy dissipation equilibrium rate.