[Paper Review] Beeping a Maximal Independent Set

Yehuda Afek, Noga Alon+4 more|arXiv (Cornell University)|Jun 1, 2012

Mobile Ad Hoc Networks20 references4 citations

TL;DR

This paper presents a distributed algorithm for computing a Maximal Independent Set (MIS) in a harsh beeping radio network model with only carrier sensing, no synchronization, and adversarial wake-up. It proves a sub-polynomial time lower bound and introduces four relaxations—known network size, wake-on-beep, sender collision detection, and synchronous clocks—each enabling polylogarithmic-time MIS computation, with the fastest at O(log²n) rounds under favorable conditions.

ABSTRACT

We consider the problem of computing a maximal independent set (MIS) in an extremely harsh broadcast model that relies only on carrier sensing. The model consists of an anonymous broadcast network in which nodes have no knowledge about the topology of the network or even an upper bound on its size. Furthermore, it is assumed that an adversary chooses at which time slot each node wakes up. At each time slot a node can either beep, that is, emit a signal, or be silent. At a particular time slot, beeping nodes receive no feedback, while silent nodes can only differentiate between none of its neighbors beeping, or at least one of its neighbors beeping. We start by proving a lower bound that shows that in this model, it is not possible to locally converge to an MIS in sub-polynomial time. We then study four different relaxations of the model which allow us to circumvent the lower bound and find an MIS in polylogarithmic time. First, we show that if a polynomial upper bound on the network size is known, it is possible to find an MIS in O(log^3 n) time. Second, if we assume sleeping nodes are awoken by neighboring beeps, then we can also find an MIS in O(log^3 n) time. Third, if in addition to this wakeup assumption we allow sender-side collision detection, that is, beeping nodes can distinguish whether at least one neighboring node is beeping concurrently or not, we can find an MIS in O(log^2 n) time. Finally, if instead we endow nodes with synchronous clocks, it is also possible to find an MIS in O(log^2 n) time.

Motivation & Objective

• To address the challenge of computing an MIS in a minimalistic beeping model with no coordination, no knowledge of network size, and adversarial node wake-up times.
• To establish fundamental limits by proving a lower bound showing sub-polynomial time convergence is impossible in the basic model.
• To explore practical relaxations of the model that enable efficient, polylogarithmic-time MIS computation.
• To demonstrate that even with minimal communication (beeps and carrier sensing), non-trivial distributed tasks like MIS can be solved efficiently under mild assumptions.
• To bridge theoretical distributed computing with real-world applications in wireless sensor networks and biological systems like SOP selection in flies.

Proposed method

• Introduces a randomized, distributed algorithm based on Luby’s MIS algorithm, adapted to the beeping model with local estimation of node priority using a parameter k.
• Uses a hierarchical, power-of-two doubling mechanism for k to allow nodes to self-estimate their relative priority and avoid conflicts.
• Employs two types of signals: a 'Restart-Bit' to propagate k updates and a 'Competing-Bit' to detect conflicts and trigger k increases.
• Relies on high-probability concentration bounds to ensure that k values stabilize quickly and that no two neighbors enter the MIS simultaneously.
• Leverages locality arguments: if a node’s O(log n)-neighborhood behaves like a well-synchronized Luby instance, the node stabilizes quickly.
• Introduces four model relaxations: (1) known polynomial upper bound on network size, (2) wake-on-beep (nodes wake when they hear a beep), (3) sender-side collision detection, and (4) synchronous clocks, each enabling faster convergence.

Experimental results

Research questions

• RQ1 Can a distributed MIS algorithm be designed in a beeping model with only carrier sensing, no synchronization, and adversarial wake-up?
• RQ2 What are the fundamental time complexity limits of solving MIS in such a minimalistic model?
• RQ3 Which minimal assumptions (e.g., network size bound, wake-on-beep, collision detection, clocks) enable polylogarithmic-time MIS computation?
• RQ4 How does the proposed algorithm ensure correctness and convergence despite lack of global coordination and limited feedback?
• RQ5 To what extent can biological and wireless network models be abstracted using the beeping model for distributed computation?

Key findings

• A lower bound of Ω(√(n/log n)) rounds is proven for any randomized algorithm to compute an MIS in the basic beeping model with adversarial wake-up, even in the stronger radio network model with collision detection.
• With a known polynomial upper bound on network size, the algorithm computes an MIS in O(log³n) rounds with high probability.
• Under the wake-on-beep assumption (nodes wake when they hear a beep), the algorithm runs in O(log³n) rounds without sender collision detection.
• With both wake-on-beep and sender-side collision detection, the algorithm achieves O(log²n) rounds with high probability.
• With synchronous clocks (even without wake-on-beep), the algorithm computes an MIS in O(log²n) rounds with high probability.
• The algorithm is robust to node crashes under the same assumptions as in prior work, preserving correctness with high probability.