The TCP “Adaptive-Selection” Concept

Abstract

The rapidly increasing importance of wireless communications (including satellite), together with the rapid growth of high speed networks, pose new challenges to transmission control protocol (TCP). Among them, the most prominent are long round trip times (RTTs), not negligible packet error rates (PER), and very large bandwidths. To overcome them, a wide variety of TCP enhancements has been presented in the literature with different purposes and capabilities. However, as most proposals aim to address different impairments, they result optimized for specific network environments. Therefore, given the increasing level of heterogeneity of present and future networks, the choice of ldquothe bestrdquo TCP enhancement seems a quite irresolvable problem, depending on the characteristics of the specific connections. The TCP adaptive-selection concept, presented and discussed in this paper, aims to circumvent this problem by providing an alternative approach that challenges at the root the idea that only one TCP enhancement must be adopted, not only on the whole network in general, but also on the same server machine. In fact, by extending the concept that underlies adaptive coding and modulation (ACM) to transport layer, TCP adaptive-selection envisages the concurrent adoption of different TCP versions on the same server, the better to match the different impairments present on different connections. The implication of this novel approach, as well as the possible criteria to be adopted for the TCP selection, are deeply discussed in this paper, where a particular emphasis is given to the ldquodynamicrdquo TCP adaptive-selection variant. Preliminary results, referring to a simple network topology, chosen to enlighten the mechanism of the TCP adaptive-selection technique, are also provided. They are quite encouraging and justify the following remarks on feasibility and the discussion of some implementation proposals.