Dynamic size counting in population protocols

TL;DR

This paper introduces a protocol for dynamic size counting in population protocols, enabling agents to estimate population size logarithmically with high probability despite adversarial additions and removals, with efficient convergence and memory use.

Contribution

The paper establishes an equivalence between dynamic size counting and loosely-stabilizing counting, and provides a protocol with optimal convergence time, polynomial holding time, and efficient memory.

Findings

Expected convergence time is O(log n + log M)

Memory usage is O(log^2(s) + (log log n)^2) bits

Protocol adapts efficiently to population size changes

Abstract

The population protocol model describes a network of anonymous agents that interact asynchronously in pairs chosen at random. Each agent starts in the same initial state $s$. We introduce the *dynamic size counting* problem: approximately counting the number of agents in the presence of an adversary who at any time can remove any number of agents or add any number of new agents in state $s$. A valid solution requires that after each addition/removal event, resulting in population size $n$, with high probability each agent "quickly" computes the same constant-factor estimate of the value $\log_2 n$ (how quickly is called the *convergence* time), which remains the output of every agent for as long as possible (the *holding* time). Since the adversary can remove agents, the holding time is necessarily finite: even after the adversary stops altering the population, it is impossible to *stabilize* to an output that never again changes. We first show that a protocol solves the dynamic size counting problem if and only if it solves the *loosely-stabilizing counting* problem: that of estimating $\log n$ in a *fixed-size* population, but where the adversary can initialize each agent in an arbitrary state, with the same convergence time and holding time. We then show a protocol solving the loosely-stabilizing counting problem with the following guarantees: if the population size is $n$, $M$ is the largest initial estimate of $\log n$, and s is the maximum integer initially stored in any field of the agents' memory, we have expected convergence time $O(\log n + \log M)$, expected polynomial holding time, and expected memory usage of $O(\log^2 (s) + (\log \log n)^2)$ bits. Interpreted as a dynamic size counting protocol, when changing from population size $n_{prev}$ to $n_{next}$, the convergence time is $O(\log n_{next} + \log \log n_{prev})$.