Improving Online Gaming Experience Using
Location Awareness and Interaction Details
Dewan Tanvir Ahmed and Shervin Shirmohammadi
Distributed and Collaborative Virtual Environments Research Laboratory
School of Information Technology and Engineering
University of Ottawa, Ontario, Canada
{dahmed, shervin}@discover.uottawa.ca


Abstract — Latency is a key element for online game quality and
user experience. The client-server approach is a widely used
system supporting hundreds of thousands of players on a regular
basis. However, if latency among players via an intermediate
server is large, timely interaction for them could be difficult. In
this article, we present a procedure to share game states for a
group of players within their area of interaction so that players
can comply with stringent time-constraint and improve their
game experience. As players move around in a game space, so do
their virtual positions. In addition, the relative orientation of
players within an area of interaction is unpredictable which
indeed changes quite frequently. Because of these facts, we
cannot make a predefined rule set for message exchange among
players. So a message exchange plan currently working well
might have low efficiency after a while due to dynamic changes.
In our procedure, considering the importance of interaction,
relative orientation, and virtual and geographical locations, we
devise a message exchange plan that works alone in each client
machine with the local information available. Significant
performance improvements are noticed through simulations,
validating our approach.
Keywords- Networked Games; client-server; latency; lag
compensation; synchronization; performance
I. INTRODUCTION
Internet-based multiplayer online games, especially First
Person Shooter (FPS) games such as Counter-Strike: Source,
and Quake III Arena introduced over a decade ago - have
become increasingly popular and significant contributors to
Internet traffic [1]. Online games usually function in a client-
server mode where players control game clients generally
running on a personal computer (PC) that communicates with
game servers hosting individual games. Because of significant
resource requirement, game publishers rely on Internet Service
Providers (ISPs), dedicated game hosting companies, and
private individuals to host game servers.
Massively Multiplayer Online Games (MMOG) support
millions of subscribers and many of them actively participate
in games on a daily basis. To manage this large number of
players, the virtual world is typically divided into realms or
kingdoms which are the clones of the same virtual world, each
hosting several thousand registered players. To be precise,
realms are geographically distributed across the Internet. So,
players from one particular region generally participate in the
same realm. Realms are further divided into separate areas.
Each area is considered as a zone. Zones can have different
themes and different level of difficulties — in order to hold
inexperienced players advancing into the next hard level.
Normally, there is a server for each zone managing game
traffic. The communication strategy within a zone best
resembles to a multicast structure because of players’ common
interest (i.e. regular interaction) in game logic. IP multicasting
can be an ideal solution which was initially proposed for group
communication. But IP multicast is designed for hierarchical
routing and does not scale well in terms of supporting a large
number of concurrent groups. The deployment hurdles and
service provider’s aversion to provide IP multicast make
matters worse [2][3]. Current practices therefore deeply rely
on centralized architectures that cause scalability bottlenecks
(in terms of maximum number of people supported in a single
realm) and are also costly to adopt and deploy. To address this
problem, new designs are proposed attempting to include both
client and server side resources in a seamless manner to take
advantage of the P2P paradigm [2][4], with the objective of
utilizing client side recourses, quick deployment and better
service. In this paper, we assume a client-server infrastructure,
as currently this is the most widely-deployed method to
support MMOGs.
In order to support a consistent game space, each player
must keep a clone of relevant game states in his computer.
When a player performs an action or generates an event
affecting the virtual space, the update must be shared with
other players around. The amount of data required to exchange
roughly depends on population size of the interested area.
However, the capacity is bounded by at least two practical
limitations — network bandwidth and processing power
[5][6]. Latency tolerance usually varies from game to game
and is typically limited to a value between 100ms to 1000ms
depending on many factor such as game perspective (i.e. First-
person or Third-person), game genres (i.e. racing or role
playing game), and the sensitivity of the actions [7].
In this article, we devise a game state sharing mechanism
that helps players to obey time constraint of the application. In
online games, virtual and physical positions are two important
terms. The first one defines a player’s location in the game
space while the latter represents the actual position of a
player’s computer in the Internet (as a node in the network).
Both are significant and play a role in our design. We believe
that in the same interaction space, the importance of

interaction among players is non-uniform. Simply, a closer
player has higher importance than a distant one, which is
intuitive in some sense. On the other hand, symmetric and
asymmetric relationships (explained later) among players are
not fixed in a game space and change over time based on
players’ relative position and orientation. So a message
prioritization procedure should take these temporal symmetric
and asymmetric relationships into account while formulating
rules to exchange game states. Considering the importance of
interaction, relative orientation, and virtual and geographical
locations, we propose a message exchange plan that works
standalone in each client machine with the local information
available.
After giving an overview of related work in the following
section, we state the problem and objectives in Section III. In
Section IV, we introduce the algorithm to comply with the
time constraint. In Section V, a quality control mechanism is
outlined. The new message exchange plan for online games is
presented in Section VI. The performance analysis and other
relevant issues are discussed in Section VII. Finally, the paper
is concluded in Section VIII.
II. RELATED WORK
The early proposals of online games - suggested decades
ago - were purely Client-Server in nature where an update
message is sent from one player to the server who then relays
it to all other players. While peer-to-peer (P2P) approaches
have recently been proposed to either move away from a
client-server approach or to complement it, it is apparent that a
pure P2P architecture is not a viable solution by gaming
companies due to business prospect, security, and quality
concerns. Also, there are serious technical doubts about the
practicality of a purely P2P approach for MMOGs [8]. At
present, different mixed or hybrid architectures are being
proposed using peer resources, but practical deployment
hurdles are not yet fully overcome.
Consistency, responsiveness, reliability, security, and
persistency are fundamental concerns of MMOGs [9][6]. Real-
time applications like online games demand realization of an
action in time. The general phenomenon is that, when a player
interacts with other players, updated information must be sent
to all the participants. Network latency and processing delay
make the matter hard. Because of networking limitations and
traffic conditions, some of these updates might be lost or
delayed. Much research has been conducted to overcome the
networking limitations and to make more reliable systems.
Some of these studies provide receiver-initiated and
selectively-reliable transport protocols [10] that can be used to
deliver important messages with a high degree of reliability,
while others use sender-initiated approaches to deliver key
updates with guaranteed reliability [11]. The IEEE DIS
standard [12] has also been successfully used in a controlled
environment with vast resources, mostly for military
simulations.
Intra-zone communication means communication inside a
zone or within a group. The players who are close to each
other in a virtual world interact more frequently than others
and are involved in many common activities. In order to retain
consistency, there is a need to exchange messages among
themselves, and intra-zone communication becomes
necessary. For local communication inside a zone, the zone
master’s and players’ active participations are combined to
overcome the resource limitations of the master. This is a kind
of overlay-based state-sharing mechanism. As mentioned
earlier, the acceptable latency value for online games appear to
be game specific and defined by upper bounds. For example,
Fritsch et al. show that the game Everquest2 can run with a
latency up to 1250ms [13]. For this purpose, authors monitor
the time span in which players kill a certain number of
monsters, outstanding health, and magic points that players
have after the encounter. Dick et al. conduct player survey and
analyze how latency affects player performance in different
games [14]. According to this, a Role Playing Game like
Diablo II performs well when latency is around 80ms which
can be stated as the optimal, and a latency of 120ms as a
maximum tolerable value. While the client-server paradigm is
easy to deploy and is commonly used, the fact that a message
must go through an intermediary node (server) to reach its
destination can add unwanted delay. In this article, we present
a state sharing procedure that allows players to exchange game
state directly when the experienced latency via intermediate
server exceeds the threshold.
Marios et al. present an approach to support massively
multiplayer online role-playing games (MMORPGs) using a
centralized distributed architecture [15]. This approach
considers a player's locality of interest to reduce bandwidth
requirements for both game servers and clients. But from an
architectural point of view, it is simply a multiple server-based
client-server architecture where performance improvement is
flat. There is no guarantee on end-to-end delay. Here, a player
state includes a set of all other players and servers that
currently know this player. However, if a player leaves, it is
not clear what will happen to others, i.e. how will this
departure be handled? This effect of player departure is not
addressed in this architecture.
The model proposed by Hampel et al. reuses architectures
capable of exploiting the flexibility and scalability of P2P
networks [16]. One of the main drawbacks of P2P networks
for games is the lack of a central authority that can regulate
access and prevent cheating. The model in [14] overcomes that
by using a set of controller peers that can supervise each other.
This kind of redundancy can prevent cheating. The model is
based on the existing distributed hash table Pastry, which has
been extended into SCRIBE. The key issue is unbounded end-
to-end delay, which could be a problem for synchronization in
online games.
Our proposed approach works in client-server mode but
includes active participation of players when latency affects
gaming experience. Out of many players, important pairs of
players are identified considering attributes such as player’s
geographical placement, virtual position in the game space,
and their relative orientation. Earlier we have considered only
physical and virtual positions of players while developing the
message forwarding mechanism [17]. In this article, we also