Victor Rangel

Route duration modeling for mobile ad-hoc networks

In this paper, we present a model that estimates the time duration of routes formed by several intermediate nodes in mobile multi-hop ad-hoc networks. First, we analyze a 3-node route, where only the intermediate node is in movement while source and destination nodes remain static. From this case, we show how route duration is affected by the initial position of the intermediate node and the size of the region where it is located. We also consider a second case where all nodes of 3-node routes are mobile. Based on extensive analysis of these routes, we determine the PDF of route duration under two different mobility models. This PDF can be determined by either analytical or statistical methods. The main contribution of this paper is that the time duration of a route formed by N intermediate nodes can be accurately computed by considering the minimum route duration of a set of N routes of 3 nodes each. Simulation work was conducted using the NS-2 network simulator to verify the accuracy of the proposed model and to compare it with other proposals found in the literature. We show that our model is in better agreement with simulation results as compared with other models. Results from this work can be used to compute overhead signaling during route-maintenance of unicast and multicast routing protocols for mobile ad-hoc networks. Similarly, because route duration decreases with route length, this study can be used to scale the network size up/down.

GUIDE-gradient: A Guiding Algorithm for Mobile Nodes in WLAN and Ad-hoc Networks

Whereas there is a lot of work related to finding the location of users in WLAN and ad-hoc networks, guiding users in these networks remains mostly an unexplored area of research. In this paper we present the concept of node-to-node guidance and introduce a method that can be used to implement it. This method relies on the computation of a local gradient in the neighborhood of the moving node. We named this protocol GUIDE-gradient, which is a GPS-free and infrastructure-free node-to-node guiding system. In this paper we also discuss how the guiding algorithm can be generalized to node-to-node guidance in multihop ad-hoc networks.

Delivery of low rate isochronous streams over the digital video broadcasting/Digital Audio-Visual Council cable television protocol

This paper examines the capacity and performance characteristics of the digital video broadcasting (DVB)/Digital Audio-Visual Council (DAVIC) cable television protocol for the delivery of low rate isochronous streams for a cable population of up to 700 nodes. Streams (ranging from 8 to 128 kbps) suitable for timing critical services such as compressed/uncompressed voice (e.g., VoIP: G711 and G.7231), audio and low quality video, were considered in order to study the effects on channel capacity when using reservation and fixed access for the delivery of timing critical services. The analysis focuses on the,performance of the upstream channel, which is the limiting factor of community antenna television (CATV) networks and is critical in the delivery of services to individual subscribers on demand. Simulation results indicated that such streams, within the given protocol limitations, can be supported for a particular system population with trade-offs in terms of system throughput and channel utilization. Network capacity, in terms of the number of simultaneous streams supported and link utilization, is significantly affected by packet size. Analysis of the results indicated that for different streams, packet sizes and combined with header suppression, the benefits from the use of fixed access is essential for the support of timing critical services.

Performance analysis and optimisation of the digital video broadcasting/Digital Audio Visual Council cable modem protocol for the delivery of isochronous streams

This paper presents a performance analysis of the European cable communications standard ETSI EN 200 800, developed by the digital video broadcasting (DVB) project in collaboration with the Digital Audio Visual Council (DAVIC) organization. The performance is based on the medium access control (MAC) protocol of the DVB/DAVIC standard. This protocol uses, by default, a limited reservation-access mechanism for the transmission of reservation requests. It was not optimised for the delivery of isochronous streams (timing critical interactive services), as is evident from the increased risk of collision with reservation requests among stations. Optimisation of this application is becoming more significant since alternative technologies, e.g. ADSL, are becoming popular. Therefore, five enhanced reservation request mechanisms are evaluated and analysed, each of which improves significantly overall system performance for the support of timing critical interactive services such as voice over IP and streaming video. The novel application of enhanced mechanisms presented here reduces (and/or avoids) collisions. Also, a new mechanism, the unsolicited grant slot, developed by the authors, is presented. It perfectly matches the requirements of constant bit rate (CBR) components.

A mobility-based upper bound on route length in MANETs

In mobile ad-hoc networks (MANETs) routes are usually found by means of discovery packets that are injected to the network by sender nodes. Once the intended destination is reached by a discovery packet, it replies back to the sender using the same route. Upon reception of the reply message, data transfer from sender to destination can initiate. Node mobility, however, negatively affects route duration time since position changes may lead to connectivity disruptions. Furthermore, the whole process of route discovery breaks down when, due to position changes, the route followed by a discovery packet is useless by the time it reaches the destination. In this paper the conditions leading to this effect are studied and it is shown that they impose a practical limit on how long a route can be. The paper introduces a model to compute an upper bound on route length in MANETs, which is derived from the combination of a route duration model and an access delay model for multi-hop routes. The model was validated by simulations with different network settings. From this model, it was found that the node transmission range, node mobility and total per-hop delays actually define the maximum feasible number of hops in a route. To the best of the authors’ knowledge, this is a fundamental scaling problem of mobile ad-hoc networks that has not been analyzed before from a mobility-delay perspective.

PCQoS: power controlled QoS tuning for wireless ad hoc networks

Mobile ad hoc networks typically use a common transmission power approach for the discovery of routes and the transmission of data packets. In this paper we present PCQoS; a power-controlled Quality of Service (QoS) scheme for wireless ad hoc networks which builds QoS mechanisms for specific applications that wish to tradeoff better QoS performance for sub-optimal paths. PCQoS allows selected flows to modify their transmit power as a way to add and remove relay nodes from their paths in order to coarsely modify their observed application QoS performance. We present simulation results and show that PCQoS can be used to provide coarse control over traditional QoS metrics (e.g., delay, throughput). To the best of our knowledge the PCQoS protocol represents the first attempt to use variable-range transmission control as a means to provide QoS differentiation to applications in wireless ad hoc networks.

NARD: Neighbor-assisted route discovery in MANETs

Reactive routing protocols for mobile ad-hoc networks usually discover routes by disseminating control packets across the entire network; this technique is known as brute-force flooding. This paper presents NARD, which stands for neighbor-assisted route discovery protocol for mobile ad-hoc networks. In NARD, a source node floods a limited portion of the network searching not only for the destination node, but also for routing information related to other nodes (called destination-neighbors) that were near the destination node recently. Destination-neighbors can be used as anchor points where a second limited flooding takes place in search for the destination node. Because only two limited portions of the network are flooded by control packets near the source and destination nodes, NARD can significantly reduce signaling overhead due to route-discovery compared with other proposals. Simulations with NS-2 were carried out to verify the validity of our approach.

An Upper Bound on Network Size in Mobile Ad-Hoc Networks

In this paper we propose a model to compute an upper bound for the maximum network size in mobile ad-hoc networks. Our model is based on the foundation that for a unicast route to be useful to initiate data transport, it is necessary that the time required to discover the route should be shorter than the time the route remains valid. From this model, we found that the node transmission range, mobility of nodes and number of contending nodes actually define the maximum feasible number of hops in a route, and therefore the maximum network size. Our model is derived from the combination of a route duration model, that we also derive in this paper, and a delay model for multi-hop routes extended from a single-hop delay model found in the literature. We evaluate our model numerically for different network conditions. To the best of our knowledge, this is the first time that a comparison between route discovery time and route duration is analyzed in order to establish the maximum network size in ad-hoc networks. We believe this is a fundamental scaling problem of ad-hoc networks that has not been looked at before from a mobility-delay perspective.

NARD: Neighbor-Assisted Route Discovery in Wireless Ad Hoc Networks

Routing protocols for mobile ad hoc networks usually discover routes by flooding the entire network with control packets; this technique is known as blind flooding. This paper presents NARD, a neighbor-assisted route discovery protocol for wireless ad hoc networks. In NARD, a source node floods a limited portion of the network looking not only for the destination node, but also for routing information of other nodes (called neighbors) that were known to be near the destination node recently. Neighbor nodes can be used as anchor points where a second limited flooding takes place in search for the destination node. Because only a limited portion of the network is flooded by control packets near the source and destination nodes, NARD can significantly reduce the signaling overhead of route discovery compared with blind flooding techniques. Simulations with NS2 were undertaken to verify the validity of our approach.

Modeling Route Duration in Mobile Ad-Hoc Networks

An analytical model to estimate the time duration of routes formed by several intermediate nodes in MANETs is presented. Although there are works related to estimation of route duration, they only partially obtained analytical expressions. First, we approach the route duration problem by modeling a 3-node static case, source and destination (static) and one intermediate node (mobile). Then, a 3-node mobile case is analyzed. For both cases, we evaluate how long the route is valid. Finally, a K-node mobile case is presented. We conclude that the time duration of a route formed by N intermediate nodes can be accurately found from the minimum time duration of N separated 3-node routes. Simulations were developed using NS-2 to verify the proposed model. This work can be used to compute the signaling overhead during route-maintenance of routing protocols for MANETs and to adjust the maximum network size.