System and method for downlink power optimization in a partitioned wireless backhaul network with out-of-neighborhood utility evaluation
A system and method for downlink power optimization in a partitioned wireless backhaul network with out-of-neighborhood utility evaluation is disclosed. The method comprises performing initial downlink power optimization for each neighborhood independently, considering only in-neighborhood utilities, by obtaining the transmit powers of all hubs and a utility performance of all served remote backhaul modules (RBMs) in the neighborhood. Power optimization data for each neighborhood are then reported to a central processing unit for storage. Thereafter, for each neighborhood, an out-of-neighborhood utility evaluation is performed by the centralized processing unit, based on reported power optimization data from other neighborhoods, for example, by obtaining delta out-of-neighborhood sum utilities for each hub as a function of hub transmit power, by curve fitting of reported data. Power optimization for each neighborhood is then performed by optimizing both in-neighborhood sum utilities and out-of-neighborhood sum utilities, and hub transmit powers are updated accordingly.
1. A method of downlink power optimization in a partitioned fixed wireless backhaul network comprising a plurality of nodes, wherein each node comprises at least one hub module, and each hub module serves one or more remote backhaul modules (RBMs) through hub-RBM radio links, the network being partitioned into a plurality of neighborhoods, each comprising a subset of the plurality of nodes, the method comprising:
for each neighborhood independently, performing a downlink power optimization by optimizing a sum utility over the neighborhood comprising in-neighborhood utilities only;
reporting power optimization data to a centralized processing unit;
storing reported power optimization data for each neighborhood for look-up by the centralized processing unit;
in the centralized processing unit, for each neighborhood, performing a downlink power optimization by optimizing a sum utility comprising in-neighborhood utilities and out-of-neighborhood utilities,
said out-of-neighborhood utilities being obtained by an out-of-neighborhood utility evaluation of the stored power optimization data reported by other neighborhoods; and issuing hub transmit power update commands.
2. The method of claim 1 , wherein, for each neighborhood comprising m hubs, the reported power optimization data comprise sets of data for each of a plurality of different hub transmit powers, and each set comprising: a hub transmit power of hub m, a hub transmit power for each other hub of the neighborhood and a RBM utility of each served RBM.
3. The method of claim 2 , wherein the RBM utility comprises a throughput performance.
4. The method of claim 1 , wherein the reported power optimization data for all neighborhoods are stored by the centralized processing unit in a look-up table, and for the each neighborhood, the out-of-neighborhood sum utility evaluation comprises, for each hub m of the neighborhood, from the power optimization data reported from said other neighborhoods, obtaining a sum utility of out-of-neighborhood RBMs for each reported transmit power of hub m.
5. The method of claim 1 , wherein the reported power optimization data for all neighborhoods are stored by the centralized processing unit in a look-up table, and for the each neighborhood, the out-of-neighborhood utility evaluation comprises, for each hub m of the neighborhood, from the power optimization data reported from said other neighborhoods, obtaining a function comprising a delta sum utility of all out-of-neighborhood RBMs for each transmit power of hub m.
6. The method of claim 1 , wherein the reported power optimization data for all neighborhoods are stored by the centralized processing unit in a look-up table, and for the each neighborhood, the out-of-neighborhood utility evaluation comprises:
performing a look-up of reported power optimization data for each hub m and forming a secondary look-up table with reported power optimization data specific to each hub m;
from data in the secondary look-up table for hub m, computing a sum utility U mo of all RBMs in said one or more other neighborhoods for a current transmit power of hub m and computing a sum utility U mk of said RBMs in said one or more other neighborhoods for each other reported transmit power k of the hub m;
computing a delta sum utility ΔU mk =U mo −U mk for transmit power k relative to the current transmit power, thereby generating a plurality of data points (P m1 , ΔU m1 ) . . . (P mk , ΔU mk ); and
applying curve fitting to obtain a polynomial function fm that closely fits the data points ΔU m =f m (P m ).
7. The method of claim 6 , wherein for each neighborhood NBH i performing power optimization considering in-neighborhood and out-of-neighborhood utilities, comprises optimizing the hub transmit power levels so as to maximize the sum of RBM utilities using the following sum utility objective function:
Σ mεNBH i {U m ( P m )+β m ΔU m ( P m )}
where P m is the transmit power of the m-th hub, U m (P m ) is the utility of the m-th hub obtained given P m , ΔU m (P m ) is the delta out-of-neighborhood utility given P m , and P m is a tunable parameter.
8. The method of claim 6 , wherein the fitted polynomial function for each hub m provides a delta sum utility of all out-of-neighborhood RBMs for each power level option.
9. The method of claim 1 , wherein updating of hub transmit power levels after optimization is performed:
a) simultaneously in all neighborhoods,
or
b) sequentially by neighborhood.
10. The method of claim 1 , wherein initial power optimization data is estimated for each neighborhood using a pre-deployment planning tool and reported to the centralized processing unit for storage.
11. A centralized or distributed non-transitory computer readable storage medium storing reported data and storing programming instructions for execution by one or more processing units of a fixed wireless backhaul network comprising a plurality of nodes, wherein each node comprises at least one hub module, and each hub module serves one or more remote backhaul modules (RBMs) through hub-RBM radio links, said programming instructions implementing, in one or more of said processing units, the steps of claim 1 .
12. A system for downlink power optimization in a partitioned fixed wireless backhaul network comprising a plurality of nodes, wherein each node comprises at least one hub module, and each hub module serves one or more remote backhaul modules (RBMs) through hub-RBM radio links, the network being partitioned into a plurality of neighborhoods, each comprising a subset of the plurality of nodes, and the system comprising:
a centralized processing unit in communication with each hub;
a centralized or distributed computer readable storage medium storing programming instructions for execution by the centralized processing unit and/or processing units of each hub, said programming instructions implementing the steps of:
for each neighborhood independently, performing a downlink power optimization by optimizing a sum utility over the neighborhood comprising in-neighborhood utilities only; and reporting power optimization data to the centralized processing unit for storage;
data storage means for storing said reported power optimization data for each neighborhood for look-up by the centralized processing unit;
said programming instructions further implementing the steps of:
in the centralized processing unit, for each neighborhood, performing a downlink power optimization by optimizing a sum utility comprising in-neighborhood utilities and out-of-neighborhood utilities,
said out-of-neighborhood utilities being obtained by an out-of-neighborhood utility evaluation of the stored power optimization data reported by other neighborhoods; and
issuing hub transmit power update commands.
13. The system of claim 12 , wherein said data storage means stores, for each neighborhood comprising m hubs, reported power optimization data comprise sets of data for each of a plurality of different hub transmit powers, wherein each set comprising: a hub transmit power of hub m, a hub transmit power for each other hub of the neighborhood and a RBM utility of each served RBM.
14. The system of claim 12 , wherein the data storage means comprises a look-up table storing the reported power optimization data for all neighborhoods.
15. The system of claim 14 , wherein the data storage means further comprises secondary look-up tables, storing for each hub m, a sum utility of out-of-neighborhood RBMs for each reported transmit power of hub m, and a delta sum utility of all out-of-neighborhood RBMs for each transmit power of hub m relative to the current transmit power of hub m.
16. The system of claim 12 , wherein:
the data storage means comprises a look-up table storing reported power optimization data for all neighborhoods; and
the data storage means further comprises, for each neighborhood, secondary look-up tables storing for each hub m of the neighborhood, data for each hub m comprising a sum utility U mo of all RBMs in said one or more other neighborhoods for a current transmit power of hub m, a sum utility U mk of said RBMs in said one or more other neighborhoods for each other reported transmit power k of the hub m; a delta sum utility ΔU mk =U mo −U mk for transmit power k relative to the current transmit power, to provide a plurality of data points (P m1 , ΔU m1 ) . . . (P mk , ΔU mk ); and
the programming instructions further implement curve fitting to obtain a polynomial function f m that closely fits the data points ΔU m =f m (P m ).
17. The system of claim 16 , wherein said programming instructions perform power optimization considering in-neighborhood and out-of-neighborhood utilities, comprising optimizing the hub transmit power levels so as to maximize the sum of RBM utilities using the following sum utility objective function:
Σ mεNBH i {U m ( P m )+β m ΔU m ( P m )}
where P m is the transmit power of the m-th hub, U m (P m ) is the utility of the m-th hub obtained given P m , ΔU m (P m ) is the delta out-of-neighborhood utility given P m , and β m is a tunable parameter.
18. The system of claim 17 , wherein stored data for the fitted polynomial function for each hub m provides a delta sum utility of all out-of-neighborhood RBMs for each power level option.
19. The system of claim 12 , wherein said programming instructions further provide for updating of hub transmit power levels after optimization:
a) simultaneously in all neighborhoods,
or
b) sequentially by neighborhood.