Lower Bounds for Shared-Memory Leader Election under Bounded Write Contention

TL;DR

This paper establishes tight logarithmic lower bounds on the complexity of leader election in shared-memory systems with bounded write contention, highlighting fundamental limitations for randomized obstruction-free algorithms.

Contribution

It introduces new lower bounds for leader election algorithms in asynchronous shared-memory models with SWMR and multi-writer registers, revealing a trade-off between complexity and contention.

Findings

Logarithmic lower bounds on solo step complexity

Trade-off between complexity and contention

Extension to multi-writer register models

Abstract

This paper gives tight logarithmic lower bounds on the solo step complexity of leader election in an asynchronous shared-memory model with single-writer multi-reader (SWMR) registers, for randomized obstruction-free algorithms. The approach extends to lower bounds for randomized obstruction-free algorithms using multi-writer registers under bounded write concurrency, showing a trade-off between the solo step complexity of a leader election algorithm, and the worst-case contention incurred by a processor in an execution.