![](bg2.png)
2 RELATED WORKS
Cooperative caching improves the system
performance because it allows sharing and
coordinating the cached data among multiple mobile
users in the network. A variety of popular
cooperative caching strategies have already been
studied such as: Zone Cooperative scheme (Chand,
2007); proactive approach for cooperative Caching
(Kumar, 2010); Cache Data, Cache Path, Hybrid
Cache (Yin, 2004); COCA (Chow, 2004);
GroCOCA (Chow, 2004); Cluster Cooperative
Caching (Chand, 2006); COOP (Yu Du, 2005 &
2009). However, none of all these researches
provides a complete solution for cooperative
caching. Each of them has both advantages and
disadvantages. In this section, we only analyse
approaches which are used and compared with our
approach in more detail.
COCA (Chow, 2004) is proposed by Chow et al.
In this cooperative caching protocol, the mobile
node shares its cache contents with each other to
reduce the number of server requests and access
miss ratio. However, access latency is quite high
when the mobile hosts encounter a global cache hit.
Yu Du & et al proposed a cooperative caching
scheme called COOP for MANETs (Yu Du, 2005 &
2009). To improve data availability and access
performance, COOP addresses two basic problems
of cooperative caching. For cache resolution, COOP
uses the cocktail approach that consists of two basic
schemes: hop-by-hop resolution and zone-based
resolution. By using this approach, COOP discovers
data sources, which have less communication cost.
For cache management, COOP uses the inter- and
intra-category rules to minimize caching
duplications between the nodes within a same
cooperation zone. This improves the overall capacity
of cooperated caches. Disadvantage of this scheme
is the flooding, which introduces extra discovery
overhead.
Group Caching Scheme (Ting, 2007) maintains
localized caching status of one-hop neighbors for
performing the tasks of data discovery, cache
placement, and cache replacement when a data
request is received in a mobile host (MH). Each MH
and its one-hop neighbors from a group by using
“Hello” message mechanism. In order to utilize the
cache space of each MH in a group, the MHs
periodically send their caching status to its group.
Thus, when caching placement and replacement
need to be performed, the MH selects appropriate
group member to execute the caching task. In this
scheme, the biggest concern is the energy
consumption in MHs and constrain of wireless
bandwidth. Therefore, in this GroupCaching
scheme, each MH only maintains one – hop
neighbors in a group (Radhamani, 2010).
The cooperative caching scheme in our study
named MIX-GROUP. MIX-GROUP scheme is also
built basing on the favourable characteristics of the
other ones, such as it inherited the idea of marking
data item’s label when data items are cached into
MU’s local cache. This idea is proposed in COOP
scheme (Yu Du, 2005 & 2009). However, MIX-
GROUP resolved the problem of flooding that
COOP scheme as well as other ones has met. We
will present the proposition of MIX-GROUP scheme
in detail in Section 3.
3 COOPERATIVE CACHING
MODEL
3.1 Description of Model
We assume that the cooperative caching architecture
has only one server and many mobile users (MU) as
shown in Figure 1. The server is also called the base
station (BS). Each BS controls MUs in its service
zone. The communication between the BS and MUs
is wireless link. MU is the mobile devices used to
send requests to server and can move freely in one
cell or from one cell to another cell. We are
especially interested in cooperating data among
MUs in network in this model. Each MU has a
service zone to communicate with other one. The
connection among MUs is P2P wireless network.
Figure 1: Proposed Cooperative Caching Architecture-
MIXGROUP.
In this architecture, the BS acts as the server
provide data items for MUs and it can be updated
data items. MU plays the role of both the client and
the server. MU acts as the client when it sends the
requests to retrieve data items from the BS or the
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
154