Adaptive gossip-based broadcast

Abstract

Abstract

An adaptive algorithm for efficient message diffusion in unreliable environments

“…This adaptive characteristic is precisely what distinguishes our approach from previous works, which in general do not take topology and reliability aspects into account to improve performance. As we discuss in the paper, our approach is complementary to previous optimizations proposed in the literature (e.g., [12]) and could be combined with them.…”

“…The only adaptive gossip-based reliable broadcast protocol we are aware of is [12]. In this protocol, processes adjust the message rate emission to the amount of resources available (i.e., buffer size) and to the global level of congestion in the system.…”

“…Many works have investigated this problem from a probabilistic perspective (e.g., [2,4,9,10,11,12]). Probabilistic algorithms scale much better than deterministic ones and achieve high reliability.…”

Message Buffering in Epidemic Data Dissemination

“…A different idea explored in [8] requires every process to calculate the average buffer capacity among all processes it communicates with and transmit that information. When the rate is too high with respect to the average, the process reduces the rate locally.…”

“…For reducing the memory usage, mechanisms for message drop and the number of nodes which buffer a given message are determined [5,6]. Another approach is network flow control where the idea is to influence the application by regulating its rate when processes do not have enough resources and enough time to buffer [7,8]. In message stability approach, the members inform the other peers in their view about which messages they buffer.…”