Aug 2026· International Conference on Circuit, Power and Computing Technologies· pp. 1331-1336· 0 citations· 16 references
Abstract
Artificial Intelligence systems that operate with multiple agents are increasingly being used to address workflows that are complex and distributed but the performance of multi-agent systems is often limited by the fact that the underlying communication overheads can be hidden instead of performance limits. The paper is a systematic exploration of the efficiency of inter-agent communication in three popular orchestration systems: LangGraph, CrewAI, and the OpenAI Agents SDK. A controlled benchmarking environment is crafted with the same task structure to isolate delays associated with coordination, redundancy in messages, and resource usage. The analysis shows that communication patterns have a great impact on the overall system performance, and that graph-based orchestration creates a new coordination latency, whereas sequential delegation models have a quick increase in contextual payloads. On the other hand, lightweight orchestration has lower latency but less flexibility when subject to complex workflows.The main value of the work is the introduction of a communication-focused assessment framework that measures the costs of coordination without depending on the computation of the model. Moreover, the analytically validated optimized strategies, including asynchronous execution, structured message encoding, and adaptive task scheduling are offered. The experimental results reveal that there are quantifiable improvements in the reduction of latency and the efficiency of resources with the use of these strategies. The results offer useful design tips to create scalable, high-performance multi-agent systems and form the basis of future studies on communication-aware AI orchestration.
This survey presents a systematic taxonomy and technical review of dynamic orchestration strategies designed to address communication overhead, KV cache management challenges, and increased token consumption within large Language Model-based Multi-Agent Systems.
Heet Nagoriya, H. Raithatha· International Journal of Kno...· 0 citations
This systematic review synthesizes peer-reviewed studies published between 2023 and 2026 on communication-efficient networking for distributed agentic AI, multi-agent reinforcement learning and networked autonomous systems concludes that communication efficiency should be treated as a joint optimization problem involving bandwidth, latency, computation, energy and task performance.
Recent industry practice has seen the rapid emergence of agentic systems composed of heterogeneous, tool- and LLM-mediated agent components, raising practical questions about inter-agent coordination and protocol design. This paper presents an implementation-grounded comparison of the Model Context Protocol (MCP) and the Agent2Agent (A2A) protocol, from a multi-agent systems engineering perspective, using an inter-agent coordination scenario involving LLM-based agents. We evaluate an MCP-based and an A2A-based multi-agent implementation of the same software engineering task against a set of requirements derived from prior literature and discussions with industry partners, including agent discoverability, multi-part messaging, multi-turn conversations, asynchronous communication, observability, interoperability, and access control. The results evidence that MCP can support inter-agent coordination in constrained LLM-based systems through a comparatively lightweight implementation model with lower coordination complexity, although coordination concerns such as conversational state management and task lifecycle handling must be implemented explicitly at the application layer. In contrast, A2A provides richer native support for stateful, multi-turn coordination through protocol-level abstractions for tasks and lifecycle management, but this comes with substantially greater implementation and coordination complexity. Given the narrow scope of the evaluated coordination pattern, these findings are presented as design observations from an empirical experience report rather than general claims of protocol suitability or superiority across broader classes of MAS, highlighting trade-offs and how protocol abstractions shape the distribution of coordination responsibilities in contemporary agentic systems.
Ionut Predoaia, T. Vu, Konstantinos Barmpis et al.· arXiv.org· 0 citations
Multi-agent frameworks built on large language models (LLMs) routinely entangle three logically distinct concerns: who is on the team (organization), how members align (coordination), and which algorithm fuses their work (collaboration protocol). IMACS (Intelligent Multi-Agent Collaboration System) separates the three into orthogonal, independently swappable layers. Classic organizational theory (Belbin roles, Mintzberg coordination, RACI accountability) becomes executable, validated configuration, and the framework places six published collaboration algorithms behind a common interface while exposing roles, coordination, and accountability as independently configurable factors. We use this separation to conduct controlled comparisons in which organizational assignments vary while the collaboration protocol is held fixed. It also turns protocol choice into a variable that can be learned: Adaptive Org Routing, a contextual-bandit meta-protocol, selects a protocol per task under an explicit quality-cost tradeoff, outperforms every fixed protocol in a controlled study, and trains online on real benchmark and LLM-judge rewards. The ablations expose a mechanism. Accountability placement changes outcomes exactly when the protocol routes the deliverable through the accountable agent, and the winning placement flips across model families, so organizational design cannot be hard-coded; it must be revalidated, or learned, for each model binding.
Huan-Wei Chen, Xiang Song, Jian Jin et al.· arXiv.org· 2 citations
Large language model (LLM)-driven multi-agent systems typically require multiple model invocations and complex coordination during inference, and their execution strategies directly affect system accuracy, latency, and computational cost. Parallel execution provides a means to improve inference-time efficiency. From the perspective of inference-time execution, this paper models parallelism in multi-agent systems as two distinct levels of decision processes: Replica Parallelism, which explores multiple complete solution paths at the task level, and Structural Parallelism, which enables concurrent execution within a single solution path through task decomposition. However, the roles of different forms of parallelism and their interrelationships still lack systematic study in terms of unified organization and coordination. We therefore propose TIPEX, a controllable execution framework that unifies these two levels of parallelism and coordinates their roles within the inference process under a unified execution semantics while supporting systematic combinations and analyses of different parallel strategies and parameter configurations. Systematic experiments on the GAIA benchmark demonstrate that inference-time parallelism can significantly improve accuracy and reduce end-to-end latency at the cost of increased token consumption. Further analysis shows that Replica and Structural Parallelism exhibit complementary effects across task complexities, with tasks of intermediate difficulty benefiting most from their coordination, while overly aggressive parallel strategies do not necessarily yield better performance.
AgentRadio is presented, an asynchronous message-passing layer that equips coding-agent harnesses with three primitives: threads, messages, and waiting for mentions that shows the gain growing with task difficulty, consistent with mid-course correction as the underlying mechanism.
Xinxing Ren, Qianbo Zang, Ziyan Wang et al.· arXiv.org· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
