Method and apparatus for predicting mobility based on relative mobile characteristics

• US 9,743,237 B2
• Filed: 01/02/2015
• Issued: 08/22/2017
• Est. Priority Date: 01/03/2014
• Status: Active Grant

• Current Assignee: Ajou University Industry
• Original Assignee: Ajou University Industry
• Inventors: Lee, Jae Joon, Lim, Jae Sung
• Primary Examiner(s): Tran, Cong
• Application Number:
  US14/588,682
• Publication Number:
  US 20150195806A1
• Time in Patent Office: 963 Days
• Field of Search: 4554561, 4554565, 4554566, 455457, 4554042
• CPC Class Codes: G01S 5/017 Detecting state or type of motion
• CPC Class Codes: G01S 5/0289 of multiple transceivers, e.g. in ad hoc networks
• CPC Class Codes: G01S 5/0294 Trajectory determination or predictive filtering, e.g. target tracking or Kalman filtering
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: H04W 4/023 using mutual or relative location information between multiple location based services [LBS] targets or of distance thresholds
• CPC Class Codes: H04W 4/029 Location-based management or tracking services

Abstract

An apparatus and method for predicting mobility based on relative mobile characteristics are disclosed herein. The method of predicting mobility based on relative mobile characteristics includes detecting, by each of a plurality of sensor nodes of a sensor network including the plurality of sensor nodes, a change in relative location with respect to an observed node; recording, by the each of the plurality of sensor nodes, information about the detected change in relative location; generating, by the each of the plurality of sensor nodes, a mathematical model for a possibility of approach of the observed node with respect to the each of the plurality of sensor nodes using the recorded relative location change information; and predicting probabilistically, by the each of the plurality of sensor nodes, whether the observed node will approach the each of the plurality of sensor nodes using the mathematical model.

Claims

1. A method of predicting mobility based on relative mobile characteristics, comprising:

detecting, by each of a plurality of sensor nodes of a sensor network including the plurality of sensor nodes, a change in relative location with respect to an observed node;

recording, by the each of the plurality of sensor nodes, information about the detected change in relative location;

generating, by the each of the plurality of sensor nodes, a mathematical model for a possibility of approach of the observed node with respect to the each of the plurality of sensor nodes using the recorded relative location change information; and

predicting probabilistically, by the each of the plurality of sensor nodes, whether the observed node will approach the each of the plurality of sensor nodes using the mathematical model, wherein the generating comprises;

generating a mobility state of the observed node by accumulating k pieces of relative location change information of the observed node based on changes in discrete time;

generating a state diagram using combinations of a plurality of mobility states that the observed node can have; and

generating a Markov model for probabilities of transition between a plurality of mobility states within the state diagram as the mathematical model.

2. The method of claim 1, wherein the recording comprises recording the relative location change information so that information about whether the observed node is approaching or moving away from the each of the plurality of sensor nodes is included in the recorded relative location change information.

3. The method of claim 1, wherein the predicting comprises predicting whether the observed node will approach using a probability of transition that a mobility state of the observed node at a current point in time will be a state in which the observed node is approaching in next time, which belongs to a plurality of mobility states within the state diagram.

4. A method of routing for a cognitive user (CU) in a cognitive radio network, comprising:

detecting, by the CU in the cognitive radio network, a change in relative location with respect to a primary user (PU);

recording, by the CU, information about the detected change in relative location;

generating, by the CU, a mathematical model for a possibility of approach of the PU using the recorded relative location change information;

predicting, by the CU, a possibility of approach of the PU using the mathematical model; and

setting up, by the CU, a routing path for a target node based on the predicted possibility of approach of the PU, wherein the generating comprises;

generating a mobility state of the PU by accumulating k pieces of relative location change information of the PU based on changes in discrete time;

generating a state diagram using combinations of a plurality of mobility states that the PU can have; and

generating a Markov model for probabilities of transition between a plurality of mobility states within the state diagram as the mathematical model.

5. The method of claim 4, wherein the predicting comprises predicting whether the CU will approach using a probability of transition that a mobility state of the CU at a current point in time will be a state in which the CU is approaching in next time, which belongs to a plurality of mobility states within the state diagram.

6. The method of claim 4, wherein the generating comprises generating an exponentially weighted moving average of a relative approaching speed of the PU as the mathematical model.

7. The method of claim 4, wherein the setting up comprises:

calculating communication costs, including the predicted possibility of approach of the PU, for paths via each of nodes adjacent to the CU; and

selecting one of the paths via each of nodes adjacent to the CU as the routing path based on the calculated communication costs.

8. The method of claim 4, further comprising:

broadcasting, by the CU to an adjacent CU, the predicted possibility of approach of the PU; and

receiving a broadcast message, including a possibility of approach of the PU with respect to the adjacent CU, from the adjacent CU, the possibility of approach of the PU with respect to the adjacent CU being predicted by the adjacent CU;

wherein the setting up comprises setting up the routing path with respect to the target node based on the predicted possibility of approach of the PU and the received possibility of approach of the PU with respect to the adjacent CU.

9. An apparatus for predicting mobility based on relative mobile characteristics, comprising:

a processor configured to detect a change in relative location between an observed node and a sensor node within a sensor network; and

a memory device configured to record information about the detected change in relative location;

wherein the processor is further configured to;

generate a mathematical model for a possibility of approach of the observed node with respect to the sensor node using the relative location change information recorded in the memory device;

generate a mobility state of the observed node by accumulating k pieces of relative location change information of the observed node based on changes in discrete time;

generate a state diagram using combinations of a plurality of mobility states that the observed node can have;

generate a Markov model for probabilities of transition between a plurality of mobility states within the state diagram as the mathematical model; and

predict probabilistically whether the observed node will approach the sensor node using the mathematical model.

10. The apparatus of claim 9, wherein the processor is further configured to:

predict whether the observed node will approach using a probability of transition that a mobility state of the observed node at a current point in time will be a state in which the observed node is approaching in next time, which belongs to a plurality of mobility states within the state diagram.