Information Technology Reference
In-Depth Information
TaBLe 3.2
Traditional Approach for H.245 Message Encoding
Input
Process
output
M1
Encoding
A
M2
Encoding
B
M2
Encoding
B
M3
Encoding
C
M1
Encoding
A
M1
Encoding
A
Source:
Tso, F. P., “WTS” Conference of IEEE, 2008. With permission.
TaBLe 3.3
Suggested Table Lookup Approach for Efficient H.245 Message Encoding
Input
Process
output
M1
Encoding + Reuse
A
M2
Encoding + Reuse
B
M2
Reuse
B
M3
Encoding + Reuse
C
M1
Reuse
A
M1
Reuse
A
Source:
Tso, F. P., “WTS” Conference of IEEE, 2008. With permission.
individually. Table 3.3 illustrates our suggestion of an improved approach
for H.245 message encoding in which the reuse of previous encoded bit
streams is deployed along with a minor update by dynamic compilation
process. Comparing Tables 3.2 and 3.3, we get exactly the same output but
at least three extra encoding processes can be skipped. So this leads to a
saving of at least 50% compilation time and thus provides extra free system
resources and time for better quality conversation in the post-setup phase in
this example.
3.4.3 implementation Details of Lookup Table-Based
Message encoding
Basing on the above discussions and suggestions, here we present the tech-
nical details of the lookup table-based message processing approach. An
algorithmic representation is shown in Figure 3.12 and its process flow is
illustrated diagrammatically in Figure 3.13. Basically this algorithm con-
sists of two major functions. In the first process, the system is initialized
and then a set of messages that is, M(m) are (pre)compiled and followed
by storing the return data (bit-streams) into a table. Here we defined M(m)
as a set of frequently used messages. Each time, when the system comes
