Andrew T. Campbell 
Department of Electrical Engineering and
Center for Telecommunications Research
Columbia University, New York, NY 10027-6699
http://comet.ctr.columbia.edu/~campbell
http://comet.ctr.columbia.edu/wireless

Next generation wireless communications system will be 
required to support the seamless delivery of voice, video and 
data with high quality. Delivering hard Quality of Service 
(QOS) guarantees in the wireless domain is complex due to 
large-scale mobility requirements, limited radio resources 
and fluctuating network conditions. To address this challenge 
we are developing a QOS-aware middleware platform called 
mobiware which contains the complexity of supporting 
multimedia applications operating over wireline and wireless 
ATM networks. Mobiware is a software middleware platform 
based on xbind, CORBA and Java technology that is designed to 
operate between the application and radio-link layers of 
future wireless media systems. Mobiware provides value-added 
QOS support by allowing mobile multimedia applications to 
operate transparently during handoff and periods of 
persistent QOS fluctuation.

The realization of end-to-end QOS control and the 
exploitation of scalable flows is achieved in mobiware 
through; i) resource binding between mobile devices, base 
stations and ATM switches; and ii) provision of a set of 
QOS-aware adaptive algorithms. 

These adaptive algorithms operate in unison under the control 
of mobiaware:

o QOS controlled handoff, provides signalling for 	  
handoff which exploits the use of: i) soft-state and 	  
hard-state to represent flows; ii) aggregation of flows     
to and from mobile devices; and iii) routing and QOS     
renegotiation anchor points to limit the impact of     
small-scale mobility on the wider fixed network; 
  
o adaptive network service, provides hard QOS guarantees to   
base layers (BL)   and soft QOS guarantees to enhancement   
layers (viz. E1 and E2) of multimedia flows based on the   
availability of resources in the wireless environment; and
  
o active transport, supports the transfer of multi-layer   
flows through the provision of a QOS-based API and a set   
active transport objects (e.g., media scaling).  Active
transport objects are dynamically dispatched to the mobile   
devices, base stations or ATM switches to support   
valued-added QOS at strategic nodes.