1 WP2.4 System performance employing Handoff techniques to overcome platform movement CAPANINA Virtual Meeting – July 2005 Konstantinos Katzis University.

1

WP2.4 System performance employing Handoff

techniques to overcome platform movement

CAPANINA Virtual Meeting – July 2005

Konstantinos KatzisUniversity of York

Department of Electronics

2

Overview

Intra HAP HandoffWhy Handoff?Footprints and Cells Mobility Models

Six Degrees of FreedomRandom Walk and Reflection

Cell Overlap

Immediate Handoff SimulationDesign Parameters Model ImplementationImmediate Handoff ResultsDiscussion of Results

Improved Handoff ScenarioRandom Acceptance Factor (RAF)Forced Blocked Limit (FBL)

3

Why Handoff?

Intra HAP Handoff is required to cope with:Station Keeping

HAP movements contribute on call-dropping.Intra HAP Handoff can reduce call-dropping.

User MovementsMoving “targets” such as trains travelling from cell-to-cell. To ensure continuity in service we need to perform Handoffs.

Traffic Control ManagementIntra HAP handoff can be used to redirect users connected to a highly populated cell to nearby cells in order to minimise blocking levels and ensure fairness in the system.

We will be looking at Station Keeping

4

Footprints and Cells

Footprint “Moving”

Cell (Service Area)“Fixed”

3

2

7

1

4

5

6

3

2

5

4

6

7

Footprint

Cell (Service Area)1

Cell Boundaries are based on the closest boresight centre

5

Mobility Models

Drift - x Drift - y Drift - z

Roll YawPitchx

y

z

6 Degrees of Freedom

6

Mobility Model

3

2

7

1

4

5

6

Assuming a 20km/h rotation

eg:Rotation Effect

90 Degrees Rotation

7

Mobility Models

Random WalkHAP moves at any direction at a predefined speed. HAP position is maintain within the cylindrical boundaries defined in HeliNet.

ReflectionHAP moves from current location to new location at a preddefined speed. The direction and the distance appart of the new location is randomly selected. HAP position is maintain within the cylindrical boundaries defined in HeliNet.

Start

End

Start

End

8

Cell Overlap

Cell OverlapCells on the ground overlap each other.Overlapping occurs because of the way the power decreases away from the boresight of the antenna.

Areas Formed Areas are formed on a circular footprint due to

the overlapping with other cell Overlap is maintained between three cells

Channel Allocation

Cell overlap improve use of channel allocation

Users in area B and C can be assigned channels from two or three Base stations respectively.

Intra-cell handoff performance can improve utilising cell overlap

9

Immediate Handoff Simulation

Simulation Parameters19 Cells of 3.15km radiusUniformly randomly distributed users within a radius of 9kmUsers sent messages on the uplink to identify the most appropriate beams based on received power signal, CIR levels or Traffic conditionsHandoff is centrally managed at the HAP802.16 TDMA/TDM structure is assumed to be used to allow users to connect to more than one cell simultaneously. All HAP movements are restricted within the HeliNet based model.HAP does not employ any antenna steering mechanismCPE employ antenna steering mechanism

10

Immediate Handoff Simulation

X-axis drift ResultsHandoff and Dropping probability performance

11

Immediate Handoff Simulation

X-axis drift ResultsHandoff and Dropping probability wrt distance from centre

Speed fixed at : 100km/h

12

Immediate Handoff Simulation

X-axis drift ResultsBlocking Probability

13

Immediate Handoff Simulation

Reflection, Rotation and Random Walk ResultsHandoff and Dropping probability performance

No Overlap

14

Immediate Handoff Simulation

Reflection, Rotation and Random Walk ResultsHandoff and Dropping probability wrt distance from centre

No OverlapSpeed fixed at : 100km/h

15

Immediate Handoff Simulation

Reflection, Rotation and Random Walk ResultsHandoff and Dropping probability performance

With Overlap

16

Immediate Handoff Simulation

Reflection, Rotation and Random Walk ResultsHandoff and Dropping probability wrt distance from centre

With Overlap

17

Immediate Handoff Simulation

Discussion of ResultsUsers at the edge of the cell experience increased number of handoffs. They therefore have an increased probability of being dropped.

Cell Overlap improved QoS by reducing blocking and dropping probability

Dropping probabiltiy in the case of no overlap seems to increase and levels up whereas for the case of overlap it is constantly increasing.

Blocking probability decreases at high speed in the case of no overlap due to the increased number of dropped users

Random Walk causes less Handoffs than other movements tried with the most extreme being the reflection

18

Improved Handoff Scenario

Random Acceptance Factor (RAF)Use cell overlap effectivelyEnsure uniform QoS in terms of blocking probability and bits per connection independently from the position of the userThere is a connection between the dropping and blocking probability that needs to be investigated

Forced Blocked Limit (FBL)Connection Admission Control (CAC) based on channel availability in a given cell radius of that of the new user. This will indirectly reserve channels for handing off users.The blocking is expected to increase but in a uniform manner while the dropping should become minimal.

19

Questions?

Konstantinos [email protected]