Base station wide beam codebook design
A method, an electronic device with a memory configured to store a hierarchical codebook with a processor operably connected to the memory to perform the method, and a non-transitory computer-readable medium storing instructions that when executed by a processor of an electronic device cause the electronic device to perform the method. The method identifies input data including at least one of an array size, an antenna element spacing, a phase-shifter resolution, a specified coverage region, or an antenna element pattern. The input data is processed by putting an initial beam through a non-decreasing concave utility function which uses a cyclic coordinate descent algorithm that generates a wide beam meeting one or more design specifications. A codebook is produced that includes the wide beam.
1. A method comprising:
identifying input data including at least one of an array size, an antenna element spacing, a phase-shifter resolution, a specified coverage region, or an antenna element pattern;
processing the input data and an initial beam through a non-decreasing concave utility function using a cyclic coordinate descent algorithm to generate a wide beam meeting one or more design specifications; and
producing a codebook including the wide beam.
2. The method of claim 1 , wherein processing the input data to generate the wide beam further comprises:
maximizing the non-decreasing concave utility function with multiple random initial beamforming weights, wherein the non-decreasing concave utility function has a gradient or a sub gradient and wherein the cyclic coordinate decent algorithm sequentially updates each beamforming weight until convergence.
3. The method of claim 2 , wherein maximizing the non-decreasing concave utility function further comprises:
using angular directions from a specified coverage region, with an array response at direction a(θ, ϕ), beamforming weights of the wide beam as w, beam gain pattern is P(θ, ϕ)=p(θ, ϕ)w H a(θ, ϕ)a(θ, ϕ) H w; and
identifying an equation:
max
w
∑
(
θ
,
ϕ
)
∈
C
f
(
p
(
θ
,
ϕ
)
w
H
a
(
θ
,
ϕ
)
a
(
θ
,
ϕ
)
H
w
)
,
wherein C is an angular coverage region, a(θ, ϕ) is the array response, p(θ, ϕ) is the antenna element pattern, and ƒ(x) is the non-decreasing concave utility function.
4. The method of claim 1 , wherein processing the input data to generate the wide beam further comprises one or more of:
selecting the wide beam as having a largest minimal gain over the specified coverage region, wherein the largest minimal gain meets the one or more design specifications;
varying the input data; or
choosing a design requirement comprising one of more of mean gain and mean data rate.
5. The method of claim 1 , further comprising generating the wide beam to cover a region comprising combinations of narrow beams by at least one of:
shaping the wide beam based on location of a main lobe for each of the narrow beams;
shaping the wide beam based on a contour of a composite radiation pattern of the narrow beams; or
shaping the wide beam based on one of three coverage regions centered at a direction (θ c , ϕ c ) and represented by:
Diamond: |θ−θ c |+|ϕ−ϕ c |≤d,
Circle: √{square root over ((θ−θ c ) 2 +(ϕ−ϕ c ) 2 )}≤ d , or
Square: max(|θ−θ c |,|ϕ−ϕ c |)≤ d,
wherein d is a parameter used to adjust beamwidth; and
wherein the generated wide beam comprises a beamforming vector and a beam pattern.
6. The method of claim 1 , wherein:
the one or more design specifications include at least one of a peak gain, a half-power beamwidth (HPBW), or lack of coverage holes, and
the codebook comprises a plurality of wide beams of various shapes and can cover a different number of narrow beams.
7. The method of claim 1 , further comprising using the codebook to at least one of:
depend on the size of the wide beam by applying a smaller size wide beam for a cell-center area and a larger size wide beam for a cell-edge area, wherein size is a function of a number of narrow beams covered by the generated wide beam; or
favor a lesser wide beam switching frequency to support at least one mobile user equipment (UE).
8. An electronic device comprising:
a memory configured to store a hierarchical codebook; and
a processor operably connected to the memory, the processor configured to:
identify input data including at least one of an array size, an antenna element spacing, a phase-shifter resolution, a specified coverage region, or an antenna element pattern;
process the input data and an initial beam through a non-decreasing concave utility function using a cyclic coordinate descent algorithm to generate a wide beam meeting one or more design specifications; and
produce a codebook including the wide beam.
9. The electronic device of claim 8 , wherein to process the input data to generate the wide beam, the processor is further configured to maximize the non-decreasing concave utility function with multiple random initial beamforming weights, wherein the non-decreasing concave utility function has a gradient or a sub gradient and wherein the cyclic coordinate decent algorithm sequentially updates each beamforming weight until convergence.
10. The electronic device of claim 9 , wherein to maximize the non-decreasing concave utility function, the processor is further configured to:
use angular directions from a specified coverage region, with an array response at direction a(θ, ϕ), beamforming weights of the wide beam as w, beam gain pattern is P(θ, ϕ)=p(θ, ϕ)w H a(θ, ϕ)a(θ, ϕ) H w; and
identify an equation:
max
w
∑
(
θ
,
ϕ
)
∈
C
f
(
p
(
θ
,
ϕ
)
w
H
a
(
θ
,
ϕ
)
a
(
θ
,
ϕ
)
H
w
)
,
wherein C is an angular coverage region, a(θ, ϕ) is the array response, p(θ, ϕ) is the antenna element pattern, and ƒ(x) is the non-decreasing concave utility function.
11. The electronic device of claim 8 , wherein to process the input data to generate the wide beam, the processor is further configured to one or more of:
selecting the wide beam as having a largest minimal gain over the specified coverage region, wherein the largest minimal gain meets the one or more design specifications;
varying the input data; or
choosing a design requirement comprising one of more of mean gain and mean data rate.
12. The electronic device of claim 8 , wherein the processor is further configured to:
generate the wide beam to cover a region comprising combinations of narrow beams by at least one of:
shape the wide beam based on location of a main lobe for each of the narrow beams;
shape the wide beam based on a contour of a composite radiation pattern of the narrow beams; or
shape the wide beam based on one of three coverage regions centered at a direction (θ c , ϕ c ) and represented by:
Diamond: |θ−θ c |+|ϕ−ϕ c |≤d,
Circle: √{square root over ((θ−θ c ) 2 +(ϕ−ϕ c ) 2 )}≤ d , or
Square: max(|θ−θ c |,|ϕ−ϕ c |)≤ d,
wherein d is a parameter used to adjust beamwidth; and
wherein the generated wide beam comprises a beamforming vector and a beam pattern.
13. The electronic device of claim 8 , wherein:
the one or more design specifications include at least one of a peak gain, a half-power beamwidth (HPBW), or lack of coverage holes, and
the codebook comprises a plurality of wide beams of various shapes and can cover a different number of narrow beams.
14. The electronic device of claim 8 , wherein the processor is further configured to:
use the codebook to at least one of:
depend on the size of the wide beam by applying a smaller size wide beam for a cell-center area and a larger size wide beam for a cell-edge area, wherein size is a function of a number of narrow beams covered by the generated wide beam; or
favor a lesser wide beam switching frequency to support at least one mobile user equipment (UE).
15. A non-transitory, computer-readable medium storing instructions that, when executed by a processor of an electronic device, cause the electronic device to:
identify input data including at least one of an array size, an antenna element spacing, a phase-shifter resolution, a specified coverage region, or an antenna element pattern;
process the input data and an initial beam through a non-decreasing concave utility function using a cyclic coordinate descent algorithm to generate a wide beam meeting one or more design specifications; and
produce a codebook including the wide beam.
16. The non-transitory, computer-readable medium of claim 15 , wherein the instructions to process the input data to generate the wide beam further comprise instructions that, when executed by the processor, cause the electronic device to maximize the non-decreasing concave utility function with multiple random initial beamforming weights, wherein the non-decreasing concave utility function has a gradient or a sub gradient and wherein the cyclic coordinate decent algorithm sequentially updates each beamforming weight until convergence.
17. The non-transitory, computer-readable medium of claim 16 , wherein the instructions to maximize the non-decreasing concave utility function further comprise instructions that, when executed by the processor, cause the electronic device to:
use angular directions from a specified coverage region, with an array response at direction a(θ, ϕ), beamforming weights of the wide beam as w, beam gain pattern is P(θ, ϕ)=p(θ, ϕ)w H a(θ, ϕ)a(θ, ϕ) H w; and
identify an equation:
max
w
∑
(
θ
,
ϕ
)
∈
C
f
(
p
(
θ
,
ϕ
)
w
H
a
(
θ
,
ϕ
)
a
(
θ
,
ϕ
)
H
w
)
,
wherein C is an angular coverage region, a(θ, ϕ) is the array response, p(θ, ϕ) is the antenna element pattern, and ƒ(x) is the non-decreasing concave utility function.
18. The non-transitory, computer-readable medium of claim 15 , wherein the instructions to process the input data to generate the wide beam further comprise instructions that, when executed by the processor, cause the electronic device one or more of:
select the wide beam as having a largest minimal gain over the specified coverage region, wherein the largest minimal gain meets the one or more design specifications;
vary the input data; or
choose a design requirement comprising one of more of mean gain and mean data rate.
19. The non-transitory, computer-readable medium of claim 15 , wherein further comprising instructions that, when executed by the processor, cause the electronic device to:
generate the wide beam to cover a region comprising combinations of narrow beams by at least one of:
shape the wide beam based on location of a main lobe for each of the narrow beams;
shape the wide beam based on a contour of a composite radiation pattern of the narrow beams; or
shape the wide beam based on one of three coverage regions centered at a direction (θ c , ϕ c ) and represented by:
Diamond: |θ−θ c |+|ϕ−ϕ c |≤d,
Circle: √{square root over ((θ−θ c ) 2 +(ϕ−ϕ c ) 2 )}≤ d , or
Square: max(|θ−θ c |,|ϕ−ϕ c |)≤ d,
wherein d is a parameter used to adjust beamwidth; and
wherein the generated wide beam comprises a beamforming vector and a beam pattern.
20. The non-transitory, computer-readable medium of claim 15 , wherein:
the one or more design specifications include at least one of a peak gain, a half-power beamwidth (HPBW), or lack of coverage holes, and
the codebook comprises a plurality of wide beams of various shapes and can cover a different number of narrow beams.