Timely Arbitrator Selection in P2P Virtual Environments with 3D Voronoi Diagrams Mahathir Almashor, Ibrahim Khalil Royal Melbourne Institute of Technology (RMIT) University Email: {m.almashor, ibrahim.khalil}@rmit.edu.au Abstract—Dynamically selecting game-play arbitrators re- mains a key concern in fully Peer-to-Peer Virtual Environments (P2P-VE). The lack of hierarchical structures and dedicated servers mean suitable candidates must be chosen from amongst the peer population. These selected peers are elevated to the role of momentary referees, deciding on the outcomes of interactions (e.g., combat) between adversarial peers. Accordingly, a timely selection process will aid game-play responsiveness in such time- sensitive applications. There is a need to promptly affirm the arbitrator as delays will hinder peers from initiating combat, impeding game-play. The aim is to address this singular issue, investigating a myriad of selection policies contrived within the context of 3D Voronoi Diagrams (3D-VD). Prior art utilized basic 2D varieties to spatially cluster peers and limit network traffic growth. The work presented here augments this approach with a non-spatial Z-axis and the use of unique selection algorithms. Each mechanism is discussed and extensively tested, with detailed simulation results presented herein. We thus demonstrate 3D- VD’s unique ability to deterministically appoint such on-the-fly adjudicators from localised candidate pools in a timely manner. Index Terms—Peer-to-peer Systems; Overlay Networks; Dis- tributed Virtual Environments; Applied Voronoi Diagrams. I. I NTRODUCTION The overarching aim is to work towards a fully peer-to-peer virtual environment (P2P-VE). That is, the implementation of an interactive virtual world, often characterised by massively multi-player online games (MMOG), on strictly decentralised architectures. Continuing public fascination with such net- worked applications (particularly in online games) precipitate the need from within both industry and academia to address two intertwined game-play issues: security and responsiveness. The underlying link between these two requisites are game- play arbitrators. These are the adjudicators within the VE, de- ciding the outcomes of interactions amongst peers. They fulfil the need for security, ensuring game-play integrity between competitors, and also enhance responsiveness by employing latency mitigation techniques when resolving interactions. Online games [1], have their roots in Distributed Virtual Environments (DVE) [2], itself an extension of the earlier Networked VE (NVE) [3]. Derivatives are present in myriad of other fields, such as the military, social networking and collaborative tools [4]. Our concept of a fully P2P-VE encom- passes both classifications and is driven in part by prior work in [5]. It strongly defines an underlying network architecture that is fully decentralised with regards to bandwidth, processing, management and storage requirements. This paper addresses the timely selection of game-play arbitrators in a fully decentralised VE, an issue central to VE responsiveness and security. The questions include, • How do we select suitable candidate referees from amongst the total peer population? • More crucially, how do we perform this selection without negatively impacting responsiveness? Moreover, the selection process needs to be done in a com- pletely decentralised manner, in keeping with the concept of a fully P2P-VE and has to minimize any latencies introduced during the process. In effect, we are striving to choose the right candidate first as we can ill afford multiple attempts. Group voting and result validation techniques such as those proposed in [6]–[9] incur greater overheads and are (arguably) best avoided. Thus, the goal is further refined to the selection of a single arbitrator in the shortest possible time. Accordingly, we develop several selection policies, from simple euclidean distance measures to novel ones based on Newtonian mass effect laws, taking advantage of our implied 3D domain. We investigate the implications of each of our designs and analyse our obtained simulation results. This, along with the use of 3D-VD and non-positional metrics, is the main contribution of the work contained herein. The rest of this paper is as follows: §II is a summary of key ideas and related works. §III and §IV detail our general tech- nique and the assorted selection policies. Simulation results are presented in §V and §VI concludes this paper. II. BACKGROUND Systems designed on P2P principles are prized for their resource scalability, flexibility and failure resistance. In highly dynamic and heterogeneous environments such as DVEs, this proves an ideal fit. However, there are fundamental when implementing P2P-VEs, including: • Application/System Responsiveness - DVEs, and games in particular, have an almost real-time constraint when it comes to communications. This ensures players experi- ence a smooth, seamless and immersive virtual world. • Game-play Security and Consistency - Cheat resistance is essential. Players will leave if a game or VE is perceived as easily corrupted, inconsistent and unfair. • Network Traffic Scalability - The prevailing need is the minimization of bandwidth utilization, as unoptimized P2P systems generally incur far more traffic costs. !011 IEEE International Symposium on Network Computing and Applications !78-0-76!5-448!-2/11 $26.00 © 2011 IEEE DOI 10.110!/NCA.2011.40 235 !011 IEEE International Symposium on Network Computing and Applications !78-0-76!5-448!-2/11 $26.00 © 2011 IEEE DOI 10.110!/NCA.2011.40 235