Other Free Encyclopedias » Online Encyclopedia » Encyclopedia - Featured Articles » Contributed Topics from K-O

Online Gaming - Concepts, Games Classification, Gaming Architectures, Gaming Protocols, Open Problems, Conclusions

players network time

Stef ano Ferretti, Marco Roccetti, and Paola Salomoni
University of Bologna, Bologna, Italy

Definition: Online gaming refers to playing the game over the network using a personal computer, a game console, a PDA, or a smart phone.


Computer games, whether played on a classic personal computer, on a game console, on a PDA or even on a smart phone, are among the most commercially successful applications. Today, games such as Star Wars Galaxy and Everquest command audiences from 200000 to 400000 clients in US and Europe, while NCSoft’s Lineage has approximately 4 million of users in Korea. A new report from DFC Intelligence states that revenues from the game industry will bypass the music industry within the next five years, based on projected rates of expansion for the market. This report forecasts that worldwide sales of traditional PC game, video game and portable game hardware and software are expected to increase from $23.2 billion in 2003 to $31.6 billion in 2009. In a separate report, they also estimated that the online game market would reach sales of $9.8 billion by 2009. This would result in combined interactive entertainment revenues of $41.4 billion in the same year. Those are not only rumors, as they are confirmed by the following. Informa Media forecasts that in 2010 the game market will be worth of $40 million, with the new gaming sectors of broadband, mobile and interactive TV contributing 40% of this amount. Datamonitor predicts that in 2006 only wireless gaming will generate revenues of $17.5 billion worldwide. Finally, Paul Merry at the ARC Group asserts that the overall market for mobile entertainment services will top $27 billion by 2008.

Parallel to the online game market, the online game research proceeds at a fast pace as well. Indeed, highly interactive networked multiplayer games impose rigorous demands on consistent game world modifications, responsive interaction schemes among players that permit to mitigate delays and jitters in game event transmissions, and fairness of handling players’ actions. Thus, advancements on game technologies and most of all, online game technologies, are felt as urgent to provide full support to latest kinds of entertainment services.

With this in view, in the remainder of this article we will overview some aspects of paramount importance for an effective game development to ensure compelling and exciting online game experiences. In particular, first, we present a brief discussion on game typologies and overview a simple games’ categorization that helps to identify the main technical requirements needed to ensure exciting game experiences. Then, some game architectural issues are covered, followed by a discussion on commonly exploited gaming communication protocols. Finally, we survey some important open problems that still require deeper investigation.

Games Classification

The game design is primarily affected by the characteristics and properties of the game logic. Indeed, the peculiarities of a game result in different technical needs and requirements that must be guaranteed to effectively deploy such a game over a network. Thus, taxonomy of possible games may be useful to better understand common factors and differences between games.

According to, two broad game categories may be identified: skill-and-action games (also often called fast-paced games) and strategy games. While the former category is mainly characterized on perceptual and motor skills of players (i.e., the more the player is skilled to react to external stimuli in real-time the more is likely to win the game), the latter emphasizes on cognitive efforts (i.e., real-time play is rare in strategy games that mostly require more time to let the user think and then play). Combat, sport, maze and race games are examples of skill-and-action games. Adventures, D&D, Wargames, games of chance and board games are examples of strategic games. It is easy to observe that these two main categories are very diverse and present very different requirements. In fact, the primary concern of skill-and-action game developers is responsiveness, i.e., game events produced within the game must be perceived by players in real-time. Instead, the main requirements of strategic games are reliability and game consistency i.e., the evolution of the game must be described to players in a very precise and correct way, in order to allow users to effectively plain their strategies. Needless to say, due to the best effort nature of the Internet, responsive interactions are tricky requirements in online games. This complicates the development of skill-and-action online games.

Along with this categorization, we claim that today new attention must be paid on Role-Playing Games (RPGs), which in past were typically associated to classic adventure or D&D games.

Indeed, since the first birth of video games, RPGs have truly evolved in time. Focusing on other game genres, in fact, as years go by new games have been developed characterized by always more advanced graphics, advanced human-computer interfaces and advanced ways of interaction within the game, yet the way to play these games always remains the same. Online RPGs, instead, are truly different. When deployed online, these games allow players to interact within a virtual world with thousands of other players in real-time (see Figure 1). Players form alliances and collaborate to work out complex strategies, thus living a parallel (virtual) life within the virtual game environment. Online RPGs have no ending and players influence the global game plot. Summing up, these new networked games present both skill-and-action and strategic games requirements.

Finally, due to the “never-ending” nature of RPGs, the player could be so interested in following the game evolution that it would continue to play the game while moving in real world. Hence, it is easy to envisage that in a near future new online RPGs will allow players to access the game thanks to different terminals (and with the possibility of dynamically switching among different terminals e.g., from a PC to a PDA) while continuing to play the same game session.

Gaming Architectures

Classic game consoles were typically standalone platforms able to offer incredible graphical (offline) gaming experiences. Sony’s Playstation, Microsoft’s Xbox, Sega’s Dreamcast, Nintendo’s GameCube are well known products that had a great commercial success in the game market. They integrate functionalities to perform fast sprite drawing, 3D polygons, and full motion video and pulse code modulation waveform sound. As an example, new Sony Playstation 3 incorporates a Cell Processor, a multicore chip featuring a 64-bit power processor core with multiple synergistic processor cores capable of massive floating point processing. The chip is optimized for intensive gaming applications and also excels in broadband-rich media applications, such as movies. Moreover, these game consoles typically incorporate powerful Graphic Processor Units (GPUs) for sophisticated graphics renderings. For instance, Playstation exploits powerful Nvidia GPUs while ATI GPUs are used by Microsoft and Nintendo.

As of today, the new trend is connectivity. Most of all the entertainment commercial companies are equipping their game consoles with network adapters that provide access to the Internet, and more and more online-enabled game titles are released in the gaming market. Famous examples of new online games are Everquest, Half Life II, NeverWinter Nights, Quake, Ultima Online. In this sense, the experience of Korea is a clear demonstration of the possible success of online games. Indeed, Koreans are probably the more avid online game users, thanks to an extremely wide adoption of broadband access but, most of all, thanks to the incredible success of PC Baang, places where people meet and pay to play together online.

Looking forward, we claim that wireless terminals may be the games’ real future. Thanks to their new computational, graphics, storage and network capabilities, handheld devices will become real game consoles to go. However, latency in wireless games has no easy fix, especially when the attention is focused on 3G cellular networks. In fact, while these new wireless technologies guarantee a potentially “always on” connection to the game, the problems due to the possibly limited transmission throughput (caused by bandwidth constraints and possible link outages) make the cellular environment an obscure and ambiguous partner for game developers. On the other hand, the use of Wi-Fi, Bluetooth and HomeRF technologies may provide high bandwidths but only within short ranges. Thus, access points and gateways are needed in the proximity of mobile players to provide an access to large scaled game architectures.

Nevertheless, a new scenario is emerging stating that future online players’ equipment will be very heterogeneous. People will play by means of a plethora of diverse network enabled devices (e.g., PC, game consoles, handheld devices) and bandwidth connections (e.g., dialup, DSL, fiber, Wi-Fi, Bluetooth, HomeRF, 3G cellular networks). This promise certainly boosts the popularity of online games, as games and players will appear in many flavors. However, the vast diversity of possible future client configurations gives rise to a number of new issues to address.

First, games must be developed in a portable fashion to deploy them over multiple architectures and platforms. Thus, a middleware game platform, which allows developing a game once and deploys it over multiple mobile systems, is going to become a fundamental tool. Two specific developing platforms must be mentioned. First, the Java 2 platform Micro Edition (J2ME), which represents a natural solution due to the wide diffusion of Java as a developing language. J2ME is actually exploited by many developing companies such as, for example, Motorola, Nokia (see Figure 2) and Samsung. Another solution is represented by Qualcomm’s Binary Runtime Environment for Wireless (BREW). This platform provides common APIs that application developers can use to harness hardware capabilities without having to code to the system interface. Currently, BREW is exploited by Alltel, Verizon Communications and KDDI Corp.

Another important distribution issue is concerned with the management of nodes that participate to the online game. Diverse architectural solutions may be adopted, ranging from client/server to P2P -based approaches. Needless to say, when the game is deployed over large scaled networks to support a huge number of players distributed solutions must be devised In other words, focusing on Massively Multiplayer Online Games (MMOGs), emerging approaches follow the idea that the game state should be partitioned, distributed and/or replicated across multiple game servers in order to i) reduce the number of client hosts connected to a single server, ii) eliminate central bottlenecks that typically affect classic client/server solutions, iii) alleviate the problem of network congestion and iv) augment the robustness of the system. However, if portions of the game state are kept replicated across multiple servers, synchronization schemes are required to ensure the global consistency of the redundant distributed game state.

Gaming Protocols

Once game nodes have been organized to manage the distributed game system, a key point is concerned with the protocols employed for the communications among these nodes. Indeed, it is well known that skill-and-action online games suffer from slow communications, changing network conditions and consequent high levels of delay jitters. In particular, the time elapsed since the generation of a game event by a given player and the time when such an event is perceived by all other players characterizes  the pace of evolution of the game. Researchers and gamers assert that network delays should not exceed 150 milliseconds when the game advances at a constant rate Thus, much of the focus in online multi-player games is concerned with the problem of reducing response times experienced by players.

With this in view, the employed communication protocols should not increase latencies and must take care of conveying game messages to other nodes as fast as possible. In point of this fact, it is worth mentioning that a standard game communication protocol has not yet been devised because of the different requirements of developed games.

For sure, classic completely reliable transmission protocols, such as TCP, are not appropriate for fast-paced networked games. The motivation is that much of TCP’s behavior such as “congestion control” may damage the speed of the event transmission, thus hampering the achievement of real-time constraints. Moreover, it may be that game events that arrive too late may become useless; thus, a completely reliable approach may sometimes just increase transmission delays . Hence, typically only game start-up messages are sent over TCP. Alternatively, TCP may be adopted in turn-based online games (e.g., chess games) where only a single player is allowed to perform actions in the game at a given time.

Indeed, because of the real-time nature of online multi-player games, the majority of developed game communication protocols are built over unreliable (i.e., UDP) or partially reliable communication schemes. For instance, Quake, a milestone in the online game community, adopts UDP as game events transmission protocol, similarly to MiMaze that also uses RTP to manage the game session.

Obviously, one may add functionalities by developing new protocols over UDP. As an example, new ad-hoc transport protocols may be devised such as, for example, the Game Transport Protocol (GTP) that provides facilities to devise mechanisms for session management and adaptive retransmission .

As online games involve multiparty communications (i.e., many senders, multiple receivers), group communications or multicast paradigms could be really of service during the game messages delivery. In this sense, while IP multicast is actually not widely deployed, due to various technical and administrational reasons, interesting alternative proposals are represented by multicast protocols developed at the application level. In substance, each node within the game system becomes also a node of the multicast tree. Thus, overlay distribution schemes may be dynamically built that reduce the amount of messages sent throughout the network and augment the scalability of the system.

Finally, an interesting consideration is concerned with the possibility of utilizing communication protocols able to dynamically transcode contents depending on the (possibly heterogeneous) clients’ configurations. Indeed, schemes may be devised that allow different players to receive different types of information, depending on the utilized terminal (PC, game console, iTV, PDA, smart phone) and on the user capabilities (is the user on the car? Is the user blind?) Needless to say, tuning the type (and the amount) of the transmitted information opens new ways of gaming, which enable players to connect to the game at any time, from any where, and with any device (ubiquitous gaming).

Open Problems

There are plenty of open problems that still need to be addressed in order to guarantee a full support to online multiplayer games on large scale networked. The aim of this section is to overview some research issues that require investigation and are at the basis of online games.

The need for an effective game management is mainly concerned with the optimization of the resources composing the game architecture. In this context, a typical problem amounts to the fact that the processing of unnecessary game events may cause bandwidth and computational overheads, if techniques able to reduce the amount of information exchanged within the distributed game are not utilized. In other words, in certain MMOGs, players may perceive only a subset of actions produced within the game world, depending on their potential field of sight. To surmount these kinds of problems event filtering approaches have been devised that are based on spatial considerations. In substance, the game world is divided into areas of interest and the virtual map over which the game is globally carried out is split into several small pieces. Clients subscribe to one or more areas of interests and receive information sent by all others within that area. This approach has the great benefit of reducing the amount of game events transmitted among nodes of the distributed system and should be taken into serious consideration during the development of a MMOG.

Another important field of investigation is concerned with latency hiding schemes, which aim at providing users with a higher level of interactivity by compensating transmission delays and packet losses. Most of these approaches are based on dead reckoning techniques which are typically exploited to reduce the number of game events transmitted through the network. Moreover, due to their prediction schemes, dead reckoning may also compensate game event losses.

Latency often causes another typical problem that affects online games: fairness. Simply put, due to the best effort nature of the Internet, different players that connect to the game system may perceive game state updates at different times, as soon as they deliver game events transmitted though the network. In simpler words, as different network latencies and jitters characterize the network connections of different players, it is possible that certain players are treated unfairly. Fairness schemes try to compensate and hide additional latencies perceived by disadvantaged players, thus providing all the players with the same capabilities of winning the game.


The design of novel networked games demands ever increasing technical and interdisciplinary sophistications. Exciting game experiences can be provided to a large number of players connected through diverse terminals only if smart schemes are exploited for the support of online game systems. In this paper, we overviewed some main issues that require to be fully addressed along with typical solutions. Needless to say, as each game presents its own characteristics, the employed approaches must be adapted to the peculiarities of the game so as to tune the trade-off that emerge at the design level. Finally, we claim that only deep experimental assessments may guarantee the success of the devised management schemes for the support of a particular online game.

Online Multimedia and Television - Internet-to-TV [next] [back] One of Our Own

User Comments

Your email address will be altered so spam harvesting bots can't read it easily.
Hide my email completely instead?

Cancel or