Search

A sharing community prospers when participation and contribution are both high. We suggest the two, while being related decisions every peer makes, should be given separate rational bases. Considered as such, a basic issue is the viability of club formation, which necessitates the modelling of two major sources of heterogeneity, namely, peers and shared content. This viability perspective clearly explains why rational peers contribute (or free-ride when they don't) and how their collective action determines viability as well as the size of the club formed. It also exposes another fundamental source of limitation to club formation apart from free-riding, in the community structure in terms of the relation between peers' interest (demand) and sharing (supply)., Comment: accepted in WINE2005, Hong Kong, December 15-17, 2005

File-sharing systems, like many online and traditional information sharing communities (e.g. newsgroups, BBS, forums, interest clubs), are dynamical systems in nature. As peers get in and out of the system, the information content made available by the prevailing membership varies continually in amount as well as composition, which in turn affects all peers' join/leave decisions. As a result, the dynamics of membership and information content are strongly coupled, suggesting interesting issues about growth, sustenance and stability. In this paper, we propose to study such communities with a simple statistical model of an information sharing club. Carrying their private payloads of information goods as potential supply to the club, peers join or leave on the basis of whether the information they demand is currently available. Information goods are chunked and typed, as in a file sharing system where peers contribute different files, or a forum where messages are grouped by topics or threads. Peers' demand and supply are then characterized by statistical distributions over the type domain. This model reveals interesting critical behaviour with multiple equilibria. A sharp growth threshold is derived: the club may grow towards a sustainable equilibrium only if the value of an order parameter is above the threshold, or shrink to emptiness otherwise. The order parameter is composite and comprises the peer population size, the level of their contributed supply, the club's efficiency in information search, the spread of supply and demand over the type domain, as well as the goodness of match between them., Comment: accepted in International Symposium on Computer Performance, Modeling, Measurements and Evaluation, Juan-les-Pins, France, October-2005

Fairness is an important performance criterion in all resource allocation schemes, including those in distributed computer systems. However, it is often specified only qualitatively. The quantitative measures proposed in the literature are either too specific to a particular application, or suffer from some undesirable characteristics. In this paper, we have introduced a quantitative measure called Indiex of FRairness. The index is applicable to any resource sharing or allocation problem. It is independent of the amount of the resource. The fairness index always lies between 0 and 1. This boundedness aids intuitive understanding of the fairness index. For example, a distribution algorithm with a fairness of 0.10 means that it is unfair to 90% of the users. Also, the discrimination index can be defined as 1 - fairness index., Comment: DEC Research Report TR-301, September 1984

Widespread use of computer networks and the use of varied technology for the interconnection of computers has made congestion a significant problem. In this report, we summarize our research on congestion avoidance. We compare the concept of congestion avoidance with that of congestion control. Briefly, congestion control is a recovery mechanism, while congestion avoidance is a prevention mechanism. A congestion control scheme helps the network to recover from the congestion state while a congestion avoidance scheme allows a network to operate in the region of low delay and high throughput with minimal queuing, thereby preventing it from entering the congested state in which packets are lost due to buffer shortage. A number of possible alternatives for congestion avoidance were identified. From these alternatives we selected one called the binary feedback scheme in which the network uses a single bit in the network layer header to feed back the congestion information to its users, which then increase or decrease their load to make optimal use of the resources. The concept of global optimality in a distributed system is defined in terms of efficiency and fairness such that they can be independently quantified and apply to any number of resources and users. The proposed scheme has been simulated and shown to be globally efficient, fair, responsive, convergent, robust, distributed, and configuration-independent., Comment: DEC-TR-506, reprinted in C. Partridge, Ed., "Innovations in Internetworking," published by Artech House, October 1988

In this paper, we propose that the manipulator impact control problem be approached from a stochastic optimal control perspective. The reason is that not only is such approach be able to model uncertainties in contact environment, force sensing, as well as manipulator dynamics, the controllers obtained is optimally robust in terms of performance. This result is verified by analyses and simulations.