Method of multicasting
View Patent ↗A method for multicasting data includes obtaining a set of minimum utility preferences and a utility preference for each user in a multicast group, computing a corresponding probability distπbution of different multicasting data the users are requesting, running a lottery of operating points of the set according to the corresponding probability distπbution, and multicasting the operating point determined from the lottery. For scalable layered data the method includes determining initial utility and payoff parameter for each user for a lowest data layer accommodating all group users, running a lottery for the layer using an initialized probability vector to assign the lowest layer, multicasting the lowest layer, determining new initial utility and corresponding payoff parameter for each remaining group user for a plurality of layers of data, running another lottery according to a determined mixed set probability vector to select a particular layer, and multicasting the particular layer.
1. A method for multicasting data comprising:
determining a set of minimum utility preferences for each user in a multicast group;
determining utilities of data for which the users request;
running a lottery of the minimum utility preferences according to a probability distribution λ =(λ 1 , . . . λ M ) in which M is the number of operating points, wherein probabilities λ 1 through λ M are those respectively associated with a first operating point S 1 through a last operating point S M and
∑
j
=
1
N
λ
m
=
1
and where the probabilities λ 1 through λ M represent values which are within 20% of the values that correspond to the maximum value of the arithmetic product
∏
j
=
1
N
(
∑
m
=
1
M
λ
m
u
j
,
m
-
d
j
)
wherein u j,m is the utility function of the operating point m for user j and d j is the minimum utility preference for the user j.
2. The method of claim 1 comprising the step of multicasting data responsive to the step of running the lottery.
3. The method of claim 1 comprising the step of correlating minimum utility preferences to a minimum throughput or quality of service for users brought by the operating point settings.
4. The method of claim 3 comprising the step of multicasting data responsive to the step of running the lottery.
5. The method of claim 1 comprising the step of establishing relative importance of the different data that users request or relative importance of user requests such that the utilities of data correlate to relative importance in the step of determining utilities of data.
6. The method of claim 5 comprising the step of multicasting data responsive to the step of running the lottery.
7. A method for multicasting data comprising:
determining initial utility and payoff parameter for each user in a group for a lowest layer of data that accommodates all users;
initializing a probability vector as (1, . . . , 0);
running a lottery for the layer using the probability vector to assign the lowest layer;
multicasting the lowest layer;
determining new initial utility and corresponding payoff parameter for each remaining user in the group for a plurality of layers of data when there is confirmation at least one user in the group has received the lowest layer;
determining a mixed set probability vector for the plurality of layers of data;
running another lottery according to the mixed set probability vector to select a particular layer from the plurality;
multicasting the particular layer.