1. Overview of Mobile Database Caching
Presented By:
Rooma Rathore
Rohini Prinja

2. Overview Motivation Problem definition Caching Overview
Caching Granularity
Cache Coherence
Cache Replacement
Summary
Future Work

3. Motivation Popularity of mobile database applications is increasing
Caching can improve performance
Caching in mobile DB is different
Many papers have addressed different issues in mobile caching
There are not many survey papers

4. Problem Definition Given Find Objectives Constraints
Applications running on mobile databases
Find
Efficient caching mechanisms to improve the response time
Objectives
Cache is consistent at all the times
Cache-hit ratio is high
Constraints
Low bandwidth
Frequent Disconnections
Limited battery power

5. Basic Setup of Mobile Environment
Database Servers
LAN, WAN etc
MSS
MSS
MSS
MSS
Cells
MH1
MH2
MH3
MH4
MH5
MH6

6. How mobile caching is different?
Wireless transmission is much slower
Very important to minimize transmission
Frequent disconnections by client
Server might not be able to reach all clients for updates
Clients have limited battery power
Energy efficiency is important
Uplink speed is much slower
Important to reduce uplink queries from clients

7. Caching Overview Caching mechanism characterized by
Caching Granularity
What to cache?
Cache Coherence
How to maintain consistency?
Cache Replacement
Which items to discard when new ones need to be inserted ?

8. Caching Granularity
Page based caching mechanisms not suitable in mobile environment
Object Caching
Server sends all attributes of object, which are cache by client
Attribute Caching
Only attributes requested in query are sent and cached
Hybrid Caching
Cache the attributes likely to be used in future
Semantic Caching

9. Caching Granularity - Comparison
Object
Attribute
Hybrid
Semantic
Requires more cache space
But good if all attributes are used
Saves cache space
Not very good if some attributes used in one query and other in second.
Best of object and attribute
Hard to identify, which attributes will be used
Semantic information is to be saved
New queries with partial cached data can be answered while disconnected
Very good in continuous queries, where just the remaining part can be sent to server.

10. Cache Coherence 1/3 Broadly classified into 3 categories
Stateful Server
Server knows status and contents of all clients
Client Verification
Client verifies before use of cache
Stateless Server
Server doesn’t know about client’s status or its cache contents
Mostly used in mobile environment

11. Cache Coherence 2/3 Different kinds of schemes Timestamp based
Broadcasting Timestamps
Amnesic Terminals
Signature
Adaptive
Vary transmission speed
Adaptive Invalidation with Fixed Window
Adaptive Invalidation with Adjusting Window

12. Cache Coherence 3/3 Asynchronous Stateful Validity Scoped
A Stateful scheme
Based on HLC at MSS
Validity Scoped
Suitable for Location dependent data
Bit Vector with Compression (BVC)
Grouped Bit Vector (GBVC)
Implicit Scope Information (ISI)

13. Cache Coherence- Comparison
Timestamp
Adaptive
Asynchronous
Validity - based
Periodic broadcast
IR can become too long
Long query delay
Simple
Adapt to load and other parameters
Complex
Difficult to identify threshold
A stateful scheme
Additional overhead of HLC on MSS
Invalidation messages sent to individual clients so bandwidth utilization is low.
Only suitable in location dependent data environment
Doesn’t deal with data updates on server
No periodic broadcast needed

14. Cache Replacement
Tradition Schemes like LRU, LRU-K, MRU are not very suitable
Access Frequency based
Mean – mean inter operation duration
Window – highest mean in window
EWMA – recent use has higher weight
Gain based
SAIU – Stretch Access rate Inverse Update Frequency
MinSAUD – Minimum Stretch integrated with Access rates Update frequency and cache validation Delay
Target Driven

15. Cache Replacement - Comparison
Access Freq. based
Gain Based
Target driven
Uses past access frequency to predict future
Not very good if access behavior suddenly changes
Tries to minimize stretch.
Computation is complex and consumes battery power
Assumes , access rate and update frequency of item is known or can be predicted correctly
Provides generalized cost function
Optimal results
Function can be modified to optimize different parameters
Doesn’t address what to do when more than one parameter needs to be optimized.
Computation is complex

16. Summary Why mobile caching is important?
How mobile caching is different?
Discussed papers in these areas
Caching granularity
Cache coherence
Cache replacement

17. Future Work More research is needed in area of cache replacement
Study can be done to find which combination of techniques is suitable for what applications
Study to determine which algorithms are currently being used in commercial applications.

18. References 1/3
[1] Kahol, S. Khurana, S.K.S. Gupta and P.K. Srimani, “A strategy to manage
cache consistency in a distributed mobile wireless environment”, IEEE Trans.
on Parallel and Distributed System, 12(7), pp , 2001
[2] Hong V. Leong, Antonio Si “On Adaptive Caching in Mobile Databases” April
1997, Proceedings of the 1997 ACM symposium on Applied computing
[3] Alok Madhukar, Reda Alhajj “An Adaptive Energy Efficient Cache Invalidation
Scheme for Mobile Databases” April 2006, Proceedings of the 2006 ACM
symposium on Applied computing SAC '06
[4] Ken. C.K. Lee, H.V. Leong, Antonio Si “Semantic Query Caching in Mobile
Environments”, June 2005, Proceedings of the 4th ACM international workshop
on Data engineering for wireless and mobile access
[5] Chi-Yin Chow, Hong Va Leong, Alvin T. S. Chan “Distributed Group-based
Cooperative Caching in a Mobile Broadcast Environment” May 2005,
Proceedings of the 6th international conference on Mobile data management
MDM '05
[6] Boris Y. L. Chan, Antonio Si, Hong Va Leong . “Cache Management for Mobile
Databases: Design and Evaluation.” 1998, Proceedings of the Fourteenth
International Conference on Data Engineering
[7] Jianliang Xu; Qinglong Hu; Wang-Chien Lee; Dik Lun Lee; “Performance
evaluation of an optimal cache replacement policy for wireless data
dissemination” 2004, IEEE Transactions on Knowledge and Data Engineering
[8] Quen Ren, Margret Dunham, “Semantic Caching in Mobile computing” (2001)

19. References 2/3
[9] Baihua Zheng, Dik Lun Lee, “Semantic Caching in Location-Dependent Query
Processing”, Lecture Notes in Computer Science (2001)
[10] D. Barbara, “Mobile Computing and Databases—A Survey”, IEEE Trans.
Knowledge and Data Eng., vol. 11, no. 1, Jan./Feb
[11] Kristian Kvilekval, Ambuj Singh, “SPREE: Object Prefetching for Mobile
computers”, (2004), Distributed Objects and Applications (DOA)
[12] D. Barbara, T. Imielinski, “Sleepers And workalcholics : Caching strategies in
mobile computing” , Proc. ACM SIGMOD Int’l Conf. Management of Data,
[13] A. Kahol, S. Khurana, S. K. S. Gupta, P. K. Srimani, “An Efficient Cache
Maintenance Scheme for Mobile Environment” ,(2000) International
Conference on Distributed Computing Systems
[14] G. Cao, “A Scalable Low-Latency Cache Invalidation Strategy for Mobile
Environments,” IEEE Trans. Knowledge and Data Eng., vol. 15, no. 5, Sep 03.
[15] Yeung, M.K.H.; Yu-Kwong Kwok , “Wireless cache invalidation schemes with
link adaptation and downlink traffic”, IEEE Transactions on Mobile
Computing, 2005, vol4 issue 1.
[16] S. Acharya, M. Franklin, and S. Zdonik, “Dissemination-Based Data Delivery
Using Broadcast Disks,” Personal Comm., vol. 2, no. 6, Dec

20. References 3/3
[17] J. Xu, X. Tang, D. L. Lee ,“ Performance Analysis of Location-dependent
Cache Invalidation Schemes for Mobile Environments”, 2002.
[18] H. Song; G. Cao, “Cache-miss-initiated prefetch in mobile environments”, IEEE
International Conference on Mobile Data Management, 2004 Page(s):370 – 381
[19] K.-L. Wu, P. S. Yu, and M.-S. Chen. “Energy-efficient caching for wireless
mobile computing”. In 20th International Conference on Data Engineering,
pages , Feb. 26-March
[20] Xu, Q.L. Hu, and D.L. Lee, W.-C. Lee, “SAIU: An Efficient Cache Replacement
Policy for Wireless On-Demand Broadcasts”, Proc. Ninth ACM Int'l Conf.
Information and Knowledge Management, pp , Nov. 2000
[21] Chun-Hung Yuen, J.; Chan, E.; Lam, K.-Y.; Leung, H.W. “Database and
Expert Systems Applications”, Proceedings. 11th International Workshop
on 4-8 Sept Page(s):
[22] J. Jing, A. K. Elmargarmid, S. Helal, and R. Alonso. “Bit-Sequences: An
Adaptive Cache Invalidation Method in Mobile Client/Server Environments”.
ACM/Baltzer Mobile Networks and Applications, 2(2), 1997
[23] Q.L. Hu and D. L. Lee. “Adaptive cache invalidation methods in mobile
environments”. In Proceedings of the 6th IEEE International Symposium on
High Performance Distributed Computing, August 1997.
[24] L. Yin, G. Cao, and Y. Cai. “Target-Driven Cache Replacement for Mobile
Environments”, Journal of Parallel and Distributed Computing, Vol. 65, pp.
, 2005

21. Questions ?