J. J. Garcia-Luna-Aceves

J. J. Garcia-Luna-Aceves

Auto-generated profile, based on publicly available information

J. J. Garcia-Luna-Aceves

Department of Electrical and Computer Engineering, University of Toronto, Toronto, Ontario, Canada

Research Focus & Metrics

Academic Career Length

25Years (2002-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 January 2005 in Wireless Networks
Verified
Journal Article
Open access
PDF file
LBLichun Bao
Lichun Bao
JGJ. J. Garcia-Luna-Aceves
J. J. Garcia-Luna-Aceves
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.

10 March 2011 in IEEE/ACM Transactions on Networking
Verified
Journal Article
Open access
JGJ. J. Garcia-Luna-Aceves
J. J. Garcia-Luna-Aceves
RMRolando Menchaca-Méndez
Rolando Menchaca-Méndez
A framework for integrated multicast and unicast routing in mobile ad hoc networks (MANETs) is introduced. It is based on interest-defined mesh enclaves that are connected components of a MANET spanning the sources and receivers of unicast or multicast flows. The Protocol for Routing in Interest-defined Mesh Enclaves (PRIME) is presented to implement the proposed framework for integrated routing in MANETs. PRIME establishes meshes that are activated and deactivated by the presence or absence of interest in individual destination nodes and groups and confines most of the signaling overhead within regions of interest (enclaves) in such meshes. The routes established in PRIME are shown to be free of permanent loops. Experimental results based on extensive simulations show that PRIME attains similar or better data delivery and end-to-end delays than traditional unicast and multicast routing schemes for MANETs (AODV, OLSR, ODMRP). The experiments also show that signaling in PRIME is far more scalable than the one used by traditional multicast and unicast routing protocols such as AODV, OLSR, or ODMRP.

20 January 2011 in IEEE Transactions on Wireless Communications
Verified
Journal Article
Closed access
Zheng WangZheng Wang
Zheng Wang
HSHamid R. Sadjadpour
Hamid R. Sadjadpour
JGJ. J. Garcia-Luna-Aceves
J. J. Garcia-Luna-Aceves
We present capacity and delay scaling laws for random wireless ad hoc networks under all information dissemination modalities (unicast, multicast, broadcast and anycast) when nodes are endowed with multi-packet reception (MPR) capabilities. Information dissemination modalities are modeled with an (n, m, k)-cast formulation, where n, m, and k denote the number of nodes in the network, the number of destinations for each communication group, and the actual number of communication group members that receives the information (i. e., k ≤ m ≤ n), respectively. We show that Θ(R(n)\√m/k), Θ(1/k), and Θ(R2(n)) bits per second constitute a tight bound for the throughput capacity of random wireless ad hoc networks under the protocol model when m = O(R-2(n)), Ω(k) = R-2(n)= O(m), and k = Ω(R-2(n)), respectively. R(n) denotes the receiver range which depends on the decoding complexity of the nodes. For the minimum receiver range of Θ(√(log n/n)) to guarantee network connectivity, a gain of Θ(log n) for (n, m, k)-casting is attained with MPR compared to the capacity attained when receivers can decode at most one transmission at a time in . Furthermore, we derive the capacity-delay tradeoff of (n, m, k)-casting when MPR is used. We show that the use of MPR can lead to both increased network capacity and reduced delays in wireless ad hoc networks.

1 January 2010
Verified
Conference Paper
Open access
DSDhananjay Sampath
Dhananjay Sampath
SASuchit Agarwal
Suchit Agarwal
JGJ.J. Garcia-Luna-Aceves
J.J. Garcia-Luna-Aceves
Networks based on Ethernet bridging scale poorly as bridges flood the entire network repeatedly, and several schemes have been proposed to mitigate this flooding problem, however, none have managed to eliminate flooding completely. We present Automatic Integrated Routing (AIR) as the first routing protocol that eliminates flooding by assigning prefix labels to switches and building a Distributed Hash Table (DHT). The DHT maps host identifiers to the prefix labels of the switches through which they connect to the network. Each switch is assigned a prefix label using neighbor-to-neighbor messages. Prefix labels denote the locations of the switches in the network, and the prefix labels of any two switches automatically determine one or multiple routes between them. The DHT stores the mapping between the name of a host and its network location (prefix label) in a scalable fashion, with any one switch storing only a fraction of all the mappings. In contrast, prior approaches using DHTs to resolve host names incur the communication and storage overhead introduced by an underlying link-state routing protocol. Results using packet-level traces of Internet traffic demonstrate that AIR attains performance gains of orders of magnitude over Ethernet bridging and prior DHT-based schemes.

4 August 2011 in IEEE Transactions on Mobile Computing
Verified
Journal Article
Open access
JGJ.J. Garcia-Luna-Aceves
J.J. Garcia-Luna-Aceves
RMRolando Menchaca-Méndez
Rolando Menchaca-Méndez
A cross-layer framework is introduced for the effective dissemination of real-time and elastic traffic in multihop wireless networks called Scheduling and Traffic Management in Ordered Routing Meshes (STORM). Unicast and multicast routes are established in coordination with the scheduling of transmissions and bandwidth reservations in a way that bandwidth and delay guarantees can be enforced on a per-hop and end-to-end basis. The routes established in STORM are shown to be loop-free and real-time packets forwarded along these routes are shown to have bounded end-to-end delays. Results from detailed simulation experiments show that, compared to a protocol stack consisting of 802.11 DCF for channel access, AODV or OLSR for unicast routing, and ODMRP for multicast routing, STORM attains similar or better performance for elastic traffic, and up to two orders of magnitude improvement in end-to-end delays, with twice the amount of data delivery for real-time traffic while inducing considerably less communication overhead.

7 November 2002
Verified
Conference Paper
Closed access
WZW.T. Zaumen
W.T. Zaumen
SVS. Vutukury
S. Vutukury
JGJ.J. Garcia-Luna-Aceves
J.J. Garcia-Luna-Aceves
We present a practical approach to routing and anycasting with near-optimum delays taking into account the processing loads at routers and processing elements of a computer network. To accomplish this, the minimum-delay routing problem formulated by Gallager (1977) is generalized into the problem of minimum-delay routing with load-balancing to account for processing delays in network nodes (servers and routers). Gallager's theorem for necessary and sufficient conditions for minimum-delay routing is modified to include processing delays and changes of traffic levels at network nodes. The first distributed algorithm for load-balanced anycasting and routing in computer networks is presented. This algorithm, named MIDAS, provides approximate solutions to the modified necessary and sufficient conditions for minimum-delay routing. Simulations are use to compare the performance of the new algorithm with the performance of a traditional approach to sever load balancing.