978-1-4244-4547-9/09/$26.00 ©2009 IEEE TENCON 2009
Optimizing the MPLS Support for Real Time
IPv6-Flows using MPLS-PHS Approach
Imad J. Mohamad
National Advanced IPv6 (NAv6)
Universiti Sains Malaysia (USM)
Penang, Malaysia
imad@nav6.org
Tat-Chee Wan, Faisal Y. ALzyoud, Putra Sumari
School of Computer Sciences
Universiti Sains Malaysia (USM)
Penang, Malaysia
tcwan@cs.usm.my , faisal@nav6.org , putras@cs.usm.my
Abstract—The huge coverage space of IPv6 addresses and
providing guaranteed support for the ever increasing customer
demand, results in the dealing with bigger packet header-size
compared to the payload-size especially in some real time video
and audio applications, consequently more bandwidth is wasting.
Using the advantages of MPLS with IPv6 is still one of the
migration challenges at the backbone. Payload Header
Suppression (PHS) approach is currently being used in Mobile
WiMAX networks for point to point (SS to BS) communications.
The proposed approach for this paper (named) MPLS-PHS
adapt the concept of the PHS technique mentioned above,
modified to be applicable for multihop of Label Switching Path
(LSP) of MPLS domain at which multipoint-to-multipoint
connection found (Ingresses to Egresses).The results clearly show
a dramatic increase in the data throughput for real time IPv6
flows.
Keywords-component; MPLS, QoS, IPv6, Real Time
Applications and Traffic Engineering
I. INTRODUCTION
Suppression and Compression are two closely related
techniques, which deal with packet header reduction. PHS and
Robust Header Compression (RoHC) are two well known
approaches used for header reduction in wireless and satellite
environment, targeting utilization of their limited and costly
bandwidth. This utilization is not free; most available
approaches require extra processing time and extra memory
space (software and/or hardware solutions) to accomplish this
task.
Typically VoIP (as an example) uses the encapsulation
voice/RTP/UDP/IP. When MPLS labels are added, this
becomes voice/RTP/UDP/IP/MPLS-label. Also MPLS VPNs
use label stacking, and in the simplest case of IPv4 the total
packet header at least 48 bytes, while the voice payload is
often no more than 30 bytes. When IPv6 used, the relative size
of the header in comparison to the payload is even greater [1].
This paper focuses on UDP flows of real time applications.
It is organized as follows: Section II covers the essential
backgrounds. Related work is discussed in section III. The
developed methodology is discussed in section IV. Scenarios
and Simulation results explained in section V. Conclusion
outlines stated in section VI.
II. BACKGROUNDS
A. MPLS fundamental
Ingress and Egress are the input and output doors of the
MPLS cloud (Fig. 1), where labels are pushed at the former
and popped at the later. The LSR (Label Switching Router)
located one hop before the Egress is called Penultimate. The
Penultimate pops the label instead of Egress, when this facility
is activated. Core LSRs forwards the labeled packets without
considering on their layer 3 IP headers, behaving as Transit
routers.
The 32 bits MPLS header is located between MAC header
and IP header as shown in Fig. 2. At the core of MPLS cloud,
routing is done by 20 bits MPLS-label instead of 128 bits
(IPv6) of layer 3. Label Distribution Protocol (LDP) was
designed for distribution of labels inside MPLS domain. One
of the most important services that may be offered using
MPLS in general and LDP in particular is the support of
constraint-based routing of traffic across the routed network.
Figure 1. MPLS domain.
Figure 2. MPLS shim header structure.
Label
(20 bits)
EXP (Traffic Class)
(3bits)
Stack
(1bit)
TTL
(8bits)
MAC Header MPLS Header IP header Payload
32 Bits
1