Reliable Adaptive Lightweight Multicast Protocol Ken Tang Scalable Network Technologies ktang@scalable-networks.com Katia Obraczka UC Santa Cruz katia@cse.ucsc.edu Sung-Ju Lee HP Labs sjlee@hpl.hp.com Mario Gerla UCLA gerla@cs.ucla.edu Abstract— Typical applications of mobile ad hoc networks (MANET) require group-oriented services. Digital battlefields and disaster relief operations make data dissemination and teleconferences a key application domain. Network-supported multicast is hence critical for efficient any-to-many communi- cations. However, very little work has been done on “reliable” transport multicast. We propose and evaluate Reliable Adaptive Lightweight Multicast (RALM). The design choices of RALM are motivated by lessons we learned from evaluating the performance of traditional wired reliable multicast transport protocols (in particular, SRM) in ad hoc networks. We argue that two compo- nents, reliability and congestion control, are essential in designing a reliable multicast transport protocol for MANETs. RALM addresses both reliability and congestion control. It achieves reliability by guaranteeing data delivery to troubled receivers in a round-robin fashion. RALM’s send-and-wait congestion control uses NACK feedback to adjust to congestion experienced by receivers. We show through simulations that RALM achieves perfect reliability while exhibiting low end-to-end delay and minimal control overhead compared against other protocols. I. I NTRODUCTION An ad hoc network enables wireless communications with- out any fixed infrastructure or central administration. Each node communicates with each other through packet radios. Hence, every host acts as a packet forwarder as well as a source or a destination. Because of its ease of deployment, an ad hoc network is an attractive choice for scenarios where the fixed network infrastructure is non-existent (e.g., remote locations), unusable because it is insecure (e.g., covert military operations) or unavailable due to some catastrophic events (e.g., major earthquake). The types of scenarios targeted by MANETs make group-oriented services such as data dis- semination and teleconferences a key application domain. Multicast communication is an efficient means of supporting group-oriented applications. This is especially true in mobile wireless environments where nodes are energy and bandwidth constrained. Because MANETs are particularly well suited for mission-critical applications, ad hoc network protocols must provide reliable and timely data delivery even in the presence of mobility and frequent outages. MANETs have been the subject of extensive research. De- spite the fact that reliable multicasting is vital to the success of mission critical applications, surprisingly little work has been done in this area. One of the few exceptions is the Anonymous Gossip (AG) protocol [1] that recovers from losses by having pairs of multicast members exchange information on messages they have received or lost. One potential problem with this protocol is the delay it takes for nodes to recover from losses. Reliable multicast for wired networks on the other hand, has been a very active area of research [2]. MAC and transport level protocols for wireless cellular networks have also been proposed [3], [4], [5]. One may consider applying these schemes to MANETs. We argue that the design choices underlying wired reliable multicast transport protocols are not adequate for MANETs. Ad hoc networking protocols must handle node mobility. In addition, MANETs are extremely sensitive to network load and congestion, even more so than in wired shared-medium networks because of the hidden terminal problem. Generating additional control message overhead without performing adequate congestion control will considerably degrade the performance. We propose the Reliable Adaptive Lightweight Multicast (RALM) transport protocol that favors reliability and conges- tion control over throughput. Applications that are willing to trade throughput for reliability include military convert operations and search and rescue missions. For example, an operation commander disseminating mission critical data to his troops in a covert operation is more interested in reliably delivering the commands rather than obtaining high throughput (assuming adequate throughput is obtained). In such a scenario, any data loss can be fatal to the success of the entire operation. RALM is a reliable, rate-based, congestion controlled pro- tocol that targets small group operation scenarios ranging from special military operations to civilian emergency rescue applications. When there is no packet loss, RALM sends packets at the specified application sending rate. Once a loss is detected, RALM recovers by initiating a modified send- and-wait procedure. Send-and-wait is performed with each multicast receiver that experiences losses, one at a time in a round-robin fashion. Once all receivers have up-to-date packets, RALM reverts to the application sending rate. In our previous work [6], we assumed that the multicast sources know the receiver information ahead of time and was able to use a window-based congestion control approach. In this paper, we do not make such an assumption and hence use a send-and- wait procedure. We start this study by evaluating how a “wired” reliable multicast protocol performs in MANETs. While we acknowl- edge that wired protocols were not designed for MANETs, studying the behavior of these protocols in various scenar- ios will give us insights into designing new protocols for MANETs. To this end, we evaluate the performance of the Scalable Reliable Multicast (SRM) protocol [7]. SRM is one 1054 0-7803-7802-4/03/$17.00 © 2003 IEEE