Granted Patent B2

US 7,734,805 · App. 10/413,401 · Granted Jun 8, 2010

Abstract

Claims (25)

1. A method for scheduling transmissions to a plurality of users so as to enforce a minimum throughput constraint in a communication network, comprising:

determining for each user in a given timeslot a dissatisfaction metric based on a channel quality, the dissatisfaction metric being a token count of a token counter that tracks an achieved performance relative to a target minimum throughput for each user;

computing a first weight function for each user based on a rate requested by each user, an exponential function of the dissatisfaction metric and a quality of service class for each user;

computing a second weight function for each user based on the rate requested by each user; and

scheduling at least one user in a given timeslot to be served in a next scheduled transmission based on the first and second weight functions.

2. The method of claim 1 , wherein the token counter is decremented when each user is scheduled and incremented at each user's target minimum throughput.

3. The method of claim 1 , wherein the first weight function of each user is defined by

w i 1 = DRC i 1+δ ·e aT i (n) ,

where DRCi denotes the rate requested by the ith user, δ represents a rate request exponent, a denotes a tunable parameter and Ti denotes the dissatisfaction metric for the ith user; and

the second weight function is defined by

w i 2 =DRC i .

4. The method of claim 1 , wherein the target minimum throughput is a function of at least one of a quality of service (QoS) class assigned by the network to each user and a QoS class requested by each user at the start of a data session.

5. A method for scheduling transmissions to a plurality of users so as to enforce a maximum throughput constraint in a communication network, comprising:

determining, for each user in a given timeslot a satisfaction metric based on a channel quality, the satisfaction metric being a token count of a token counter that tracks an achieved performance relative to a target maximum throughput for each user;

computing a first weight function if the determined satisfaction metric is 0 or greater;

computing a second weight function if the determined satisfaction metric is negative, the second weight function being based on a rate requested by each user, an exponential function of the dissatisfaction metric and a quality of service class for each user; and

scheduling at least one of said plurality of users in the given timeslot to be served in a next scheduled transmission based on the sign of the satisfaction metric, while enforcing maximum throughput constraints for said plurality of users.

6. The method of claim 5 , further comprising:

maintaining a token counter for each user that tracks each user's achieved performance relative to the target maximum throughput of the user.

7. The method of claim 6 , wherein the token counter for each user is incremented as a function of each user's target maximum throughput in each slot, and decremented by the number of bits transmitted when each user is scheduled.

8. The method of claim 5 , wherein said second weight function is defined by

w i =DRC i 1+δi ·e aiT i R max ,

where DRCi denotes the rate requested by the ith user, δ represents a rate request exponent, a i denotes a tunable parameter and T i R max represents the satisfaction metric for the ith user.

9. The method of claim 8 , wherein said scheduling further includes selecting a user of the plurality of users having a highest positive resultant weight function to be served during a next scheduled transmission, else idling the given timeslot of the next scheduled transmission.

10. The method of claim 5 , wherein the target maximum throughput is a function of at least one of a quality of service (QoS) class assigned by the network to each user and a QoS class requested by each user at the start of a data session.

Assignments (10)