A Strategyproof Mechanism for Scheduling Divisible Loads in Linear Networks

In this paper we augment DLT (Divisible Load Theory) with incentives such that it is beneficial for processors to report their true processing capacity and compute their assignments at full capacity. We propose a strategyproof mechanism with verification for scheduling divisible loads in linear networks with boundary load origination. The mechanism provides incentives to processors for reporting deviants. The deviants are penalized which abates their willingness to deviate in the first place. We prove that the mechanism is strategyproof and satisfies the voluntary participation condition.

• Type: Conference
• Year: 2007
• Where: IPPS
• Authors: Thomas E. Carroll, Daniel Grosu
• Keywords: Distributed And Parallel Computing, Divisible Load Theory, Divisible Loads, IPPS 2007, Mechanism Provides Incentives