BACKGROUND
What is the most promising telecommunication technology of future integrated mobile commerce and other applications: IP or ATM? Common wisdom has it that a combination of these two leading technologies will provide the most promising basis for the near future. In fact, recently, IP/ATM integration technologies have emerged as advanced concepts that are expected to provide broadband multiservice to the end users by making the best utilization of IP (e.g., switching, routing, flexibility) and of ATM (e.g., traffic engineering, QoS provisioning, adaptive MAC, speed, real-time support, dynamic resource allocation in a cost-efficient manner). The multi-protocol labels switching (MPLS) over asynchronous transfer mode (ATM) broadband networks integration and interworking (MPLS/ATM, for short) were proposed as very promising extensions to the existing IP/ATM technique. MPLS/ATM is based mainly on routing, switching, and also further developing the IP capabilities. As pointed out by Lawrence (2001), two distinct types of MPLS can be envisioned :
-
packet-based MPLS: where the label switch routers (LSRs) have full packet-handling capabilities and can examine layer 3 headers on packets;
-
switch-based MPLS: where the LSRs, known as switch-based LSRs, forward packets by means of layer 2 (L2) switching. Switch-based LSRs have little or no capability to examine layer 3 (L3) headers.
Most of the switch-based LSRs used to date have been based on L2 ATM switching. The use of switch-based MPLS on ATM is referred as MPLS/ATM. These switch-based technologies ensure the creation of broadband networks with a centralized architecture . Unfortunately, its cost is unacceptably high for the deployment of public networks and of multimedia services in the context of a mass market.
We propose a novel advanced medium access control based alternative of MPLS/ATM broadband networks integration and interworking. This alternative is based on developing of an advanced long-delay broadband QoS-oriented multifunctional medium access control (MFMAC) technology proposed by Markhasin (1996, 2001) and on a completely distributed multifunctional ATM hyperbus architecture proposed in Brandt et al. (2000, September, October). The hyperbus can support wide and global areas with long-delay signal propagation characteristics. As argued by Markhasin (2001), these advanced MAC-based technologies ensure the creation of global broadband wireless and satellite networks with a completely distributed all-IP/ATM architecture . The use of MFMAC-based MPLS on ATM will be referred to as MPLS/ATM-MFMAC .
Our main goals are:
-
the development of advanced QoS-oriented protocols for MAC ATM technology along with dynamic QoS control techniques, and
-
extending, on this base, the principles of the simple and inexpensive local areas LAN and WATM architectures up to global areas ‚ metropolitan (W-MAN), widearea (W-WAN), and global-area (S-GAN) wireless and satellite ATM networks.
In order to achieve our goals, Markhasin (1996) showed that it is necessary to overcome the following three main impediments:
-
the time barrier : concerning the degradation of long-delay MAC efficiency when the round-trip time is significantly increased;
-
the dynamic barrier : concerning the difficulty of the dynamic on-the-fly control of QoS provisioning, and
-
the economic barrier : concerning the unacceptably large cost of the wireless broadband ATM networks with a centralized architecture.
The proposed advanced ATM MAC technology is specifically designed to successfully overcome these three barriers. Our main goal is to present the fundamentals of the novel all-IP/ATM and QoS-oriented advanced ATM MAC technology and dynamic QoS control technique.
