METHOD/SYSTEM FOR CONSERVING RESOURCES DURING CONVERSATION OVER WIRELESS NETWORK TRANSPORT MEDIA

• US 20120195242A1
• Filed: 02/02/2011
• Published: 08/02/2012
• Est. Priority Date: 02/02/2011
• Status: Active Grant

• Current Assignee: College of William & Mary
• Original Assignee: College of William & Mary
• Inventors: Pyles, Andrew J., Zhou, Gang, Ren, Zhen
• Granted Patent: US 8,488,505 B2
• Time in Patent Office: Days
• US Class Current: 370/311
• CPC Class Codes: G10L 2025/783 based on threshold decision
• CPC Class Codes: H04W 52/0251 using monitoring of local events, e.g. events related to user activity
• CPC Class Codes: Y02D 30/70 in wireless communication networks

Abstract

A method and system are provided for conserving network resources such as battery power of a battery-powered communication device used to support a conversation over a wireless network transport media. Periods of silence are detected during conversation taking place on a network having controllable resources such as battery power. Using the periods of silence so-detected, future silence periods occurring on the network are then predicted. Allocation of at least a portion of the controllable resources is controlled based on the future silence periods so-predicted.

Claims

1. A method of conserving battery power when using a battery-powered communication device to support a conversation over a wireless network transport media, comprising the steps of:

providing a battery-powered communication device capable of supporting a conversation over a wireless network transport media, the communication device having a high power state (HPS) of operation and a low power state (LPS) of operation wherein battery power required for said LPS is less than battery power required for said HPS;

detecting periods of silence during at least a portion of at least one conversation supported by the communication device;

predicting, using said periods of silence so-detected, future silence periods occurring during at least one of (i) a remainder of said at least one conversation, and (ii) a future conversation supported by the communication device; and

placing the communication device in said LPS during each of said future silence periods so-predicted.

2. A method according to claim 1, wherein the conversation is packetized.

3. A method according to claim 1, wherein said portion of at least one conversation is in the form of digitized samples thereof, and wherein said step of detecting comprises the steps of:

selecting a threshold audio level; and

comparing said threshold audio level to an average audio level associated with a portion of said digitized samples, wherein one of said periods of silence is indicated each time said average audio level is less than said threshold audio level.

4. A method according to claim 1, wherein said step of predicting comprises the steps of:

generating an empirical cumulative distribution function (ECDF) using said periods of silence so-detected;

generating a computed training silence length of time α

using said ECDF;

detecting a current silence period occurring during one of (i) said remainder of said at least one conversation, and (ii) said future conversation; and

determining, using said ECDF, a conditional probability that said current silence period will last an additional length of time Δ

beyond said computed training silence length of time α

, wherein one of said future silence periods is defined by (α

+Δ

) when said conditional probability is greater than or equal to a threshold probability β

.

5. A method according to claim 4, wherein said threshold probability β

is in the range of approximately 0.25 to approximately 0.7.

6. A method according to claim 4 wherein, when said conditional probability is greater than or equal to said threshold probability β

, said step of determining is repeated until said additional length of time Δ

is maximized.

7. A method according to claim 4 wherein, when said conditional probability is less than said threshold probability β

, said method further comprises the step of setting said one of said future silence periods equal to said computed training silence length of time α

.

8. A method of conserving battery power when using a battery-powered communication device to support a conversation over a wireless network transport media, comprising the steps of:

providing a battery-powered communication device capable of supporting a conversation over a wireless network transport media, the communication device having a high power state (HPS) of operation and a low power state (LPS) of operation wherein battery power required for said LPS is less than battery power required for said HPS;

selecting a threshold audio level;

comparing said threshold audio level to an average audio level of a portion of digitized samples associated with a portion of at least one conversation supported by the communication device, wherein periods of silence are indicated each time said average audio level is less than said threshold audio level;

generating an empirical cumulative distribution function (ECDF) using said periods of silence so-indicated during said step of comparing;

generating a computed training silence length of time α

using said ECDF;

detecting a current silence period occurring during one of (i) a remainder of said at least one conversation, and (ii) a future conversation supported by the communication device;

determining, using said ECDF, a conditional probability that said current silence period will last an additional length of time Δ

beyond said computed training silence length of time α

, wherein a length of a future silence period is defined by (α

+Δ

) when said conditional probability is greater than or equal to a threshold probability β

; and

placing the communication device in said LPS during said future silence period so-defined during said step of determining.

9. A method according to claim 8, wherein the conversation is packetized.

10. A method according to claim 8, wherein said threshold probability β

is in the range of approximately 0.25 to approximately 0.7.

11. A method according to claim 8 wherein, when said conditional probability is greater than or equal to said threshold probability β

, said step of determining is repeated until said additional length of time Δ

is maximized.

12. A method according to claim 8 wherein, when said conditional probability is less than said threshold probability β

, said method further comprises the step of setting said one of said future silence periods equal to said computed training silence length of time α

.

13. A system for conserving battery power in a battery-powered communication device that supports a conversation over a wireless network transport media, comprising a controller adapted to be included in a battery-powered communication device capable of supporting a conversation over a wireless network transport media, the communication device having a high power state (HPS) of operation and a low power state (LPS) of operation wherein battery power required for said LPS is less than battery power required for said HPS,

said controller detecting periods of silence during at least a portion of at least one conversation supported by the communication device,

said controller predicting, using said periods of silence so-detected, future silence periods occurring during at least one of (i) a remainder of said at least one conversation, and (ii) a future conversation supported by the communication device, and

said controller placing the communication device in said LPS during each of said future silence periods so-predicted.

14. A system as in claim 13 wherein, when said portion of at least one conversation is in the form of digitized samples thereof, said controller detects said periods of silence by comparing a threshold audio level to an average audio level associated with a portion of said digitized samples, wherein one of said periods of silence is indicated each time said average audio level is less than said threshold audio level.

15. A system as in claim 13, wherein said controller predicts said future silence periods by

generating an empirical cumulative distribution function (ECDF) using said periods of silence so-detected,

generating a computed training silence length of time α

using said ECDF;

detecting a current silence period occurring during one of (i) said remainder of said at least one conversation, and (ii) said future conversation; and

determining, using said ECDF, a conditional probability that said current silence period will last an additional length of time Δ

beyond said computed training silence length of time α

, wherein one of said future silence periods is defined by (α

+Δ

) when said conditional probability is greater than or equal to a threshold probability β

.

16. A system as in claim 15, wherein said threshold probability β

is in the range of approximately 0.25 to approximately 0.7.

17. A system as in claim 15 wherein, when said conditional probability is greater than or equal to said threshold probability β

, said controller continues determining said conditional probability until said additional length of time Δ

is maximized.

18. A system as in claim 15 wherein, when said conditional probability is less than said threshold probability β

, said controller sets said one of said future silence periods equal to said computed training silence length of time α

.

19. A method of conserving network resources, comprising the steps of:

detecting periods of silence during at least a portion of at least one conversation adapted to take place on a network having controllable resources;

predicting, using said periods of silence so-detected, future silence periods occurring on the network; and

controlling allocation of at least a portion of the controllable resources based on said future silence periods.

20. A method according to claim 19, wherein said step of predicting comprises the steps of:

generating an empirical cumulative distribution function (ECDF) using said periods of silence so-detected;

generating a computed training silence length of time α

using said ECDF;

detecting a current silence period occurring on the network; and

determining, using said ECDF, a conditional probability that said current silence period will last an additional length of time Δ

beyond said computed training silence length of time α

, wherein one of said future silence periods is defined by (α

+Δ

) when said conditional probability is greater than or equal to a threshold probability β

.

21. A method according to claim 20, wherein said threshold probability β

is in the range of approximately 0.25 to approximately 0.7.

22. A method according to claim 20 wherein, when said conditional probability is greater than or equal to said threshold probability β

, said step of determining is repeated until said additional length of time Δ

is maximized.

23. A method according to claim 20 wherein, when said conditional probability is less than said threshold probability β

, said method further comprises the step of setting said one of said future silence periods equal to said computed training silence length of time α

.