Self-organized Spectrum Chunk Selection
Algorithm for Local Area
Deployment
Sanjay Kumar 1
Yuanye Wang 2, Nicola Marchetti 2
1Birla
Institute of Technology, Ranchi, India
skumar@bitmesra.ac.in
2Aalborg
University, Denmark
ywa @es.aau.dk, nm@es.aau.dk
Contents








2
Introduction
Motivation and Objectives
Algorithm
Scenaion Description
Performance Metrics
Evaluation Parameters
Performance Results
Conclusions
Introduction











3
The IMT-A : 4G systems with new capabilities
To be operational around the year 2015
Provide access to wide range of telecommunication services
Target peak data rates 1 Gbps in DL and 500 Mbps in UL
Focus on Local Area (LA) deployment scenarios
The Wide Area (WA) deployment uses fixed frequency reuse schemes
It needs proper network planning, such as :
Frequency reuse plans
Base station location
Controlling transmit power levels etc.
The LA of IMT-A : expected random and uncoordinated deployment
Hence netwrok planning not feasible
Also, possiblly Home Node Bs (HeNBs) share the available radio spectrum
A mechanism is needed to assign spectrum in self organized manner and improve
throughput performance
Motivation and Objectives
 Reuse 2 gives the highest throughput performance in LA deployment [11], but
needs Network planning
 A mechanism is needed to achieve nearly same performance in random
deployment
Fig. 1. Possible chunk selection with
reuse 2 in random deployment
Fig. 2. Allocation for reuse 2 with
optimal network planning
To minimize the mutual inter-cell interference among HeNBs with random and
uncoordinated deployment in order to improve throughput performance .
4
Algorithm
Steps
1.
2.
3.
4.
Initialization
Update chunk selection
Interference consideration
Chunk switch over
Figure 3: Flow chart of the Algorithm
5
Algorithm Description
Spectral bandwidth is divided in two equal size spectrum chunks (for considering
to compare with reuse 2)
Algorithm could be used for any reuse schemes
Step 1: Initialization phase





6
Random selection of a spectrum chunk
Initialization with sequence number & switch-over number
Sequence number : determines position in queue
Switch-over number : helps to avoid dead lock situation
It indicates the maximum # of times a chunk is retained
Algorithm Description (1)
Step 2: Condition to update the chunk selection


The sequence number is reduced by one in each update interval
This indicates HeNB turn to update its chunk selection
Step 3: Interference consideration for chunk selection






7
Comparison of interference levels over the chunks
The UL-RIP is used as interference measurement yard stick
UL- RIP is defined as the total power received over the spectrum chunk
The chunk selection is based on : Abs (RIP1 - RIP2 ) > P_Threshold
This ensures a sufficient amount of interference difference before decision
The threshold indicates a tradeoff between :
tolerable interference vs increased signaling requirement
Algorithm Description (2)
Step 4: Switching over to other Chunk

The chunk with minimum RIP is chosen
Main features of the Algorithm








8
Uses UL- RIP (already standardized for LTE system)
Works in a random and uncoordinated deployment
Requires very little signaling exchange
Chunks allocation in sequence (avoids complexity in interference assessment)
If # of HeNBs is very high : a long convergence time may be required
Provides a scalable solution.
Performance is upper-bounded by reuses 2 scheme
Could be used for both DL & UL
Scenario
Description
50 m
100 m
Figure 4: LA office indoor with 4 HeNBs






9
HeNB Location : center of each cell (also examined with randomized)
HeNB coverage area . 50 x 25 m (consisting of 10 rooms)
Simulation method : snap-shot based
Several thousands snap-shots are simulated
Users : uniformly distributed locations
Throughput :Mapping SINR [8]
Performance Metrics



Average Cell Throughput: Cell throughput averaged among all the simulated cells
Cell Edge User Throughput: The 5% user outage throughput , CDF value
Chunk Selection Interval: Time period for one chunk selection operation.
It contains an integer number of transmission frames.
Some simplifications are also assumed:



10
No power control
Round Robin for frequency domain scheduling
Fast fading not considered
Simulation Parameters
PARAMETERS
SETTINGS
Spectrum allocation
Access schemes
Duplexing Schemes
# UE’s per Cell
Total Transmit Power
Receiver Noise Figure
Room Size
Corridoor width
Internal walls
Shadow fading
Path Loss models
LOS
NLOS
where,
100 MHz at 3.5 GHz
DL: OFDMA, UL: SCFDMA
TDD
5~10
24dBm
7-9 dB
10x10 m
5m
light attenuation (5dB)
LOS: 3dB, NLOS: 6dB (SD)
18.7 log10 (d) + 46.8 + 20log10 (f /5.0)
20 log10 (d) + 46.4 + nw · Lw + 20log10 (f /5.0)
d = direct-line distance (m)
f = carrier frequency (GHz)
nw = # of walls between Tx & Rx
Lw = wall attenuation (dB)
11
Performance Results (1)
DL Average Cell Throughput
 Improved compared to reuse
1 and 4 schemes
 Slightly lower compared to
reuse 2 scheme
 Reuse 2 needs planning for
optimal performance.
 Proposed algorithm performs in
self- organized manner with
random and un-coordinated
deployment
Figure 5 : Comparison of DL Average
Cell Throughput
12
Performance
PerformanceResults
Results (2)
DL Cell Edge User
Throughput
 significantly higher
compared to reuse 1
scheme (200 ~ 300 %)
 Nearly same as reuse
4 scheme
 However lower than
reuse 2 scheme
Figure 6: Comparison of DL Cell Edge User
Throughput
13
Performance
PerformanceResults
Results (3)
Convergence Time
The performance is
stabilized after 10
selection intervals, with
Significant Improvement
Figure 7 : Convergence Time of the Algorithm
14
Performance
PerformanceResults
Results (4)
Randomized HeNB
Locations
 It may be difficult to
control
HeNB
locations
 Especially in home
scenarios
 Nearly the similar
performance
is
realized
With the same chunk
allocation as DL, The similar
performance
has
been
realized in UL also
15
Figure 8: DL Average Cell Throughput with
Randomized HeNB Locations
Conclusions








16
An Algorithm for self organized spectrum chunk selection has been proposed
It minimizes mutual inter-cell interference and improves the system throughput performance
Performance approaches fixed frequency reuse 2 scheme
Gives much better performance than fixed frequency 1 and 4 schemes
Suitable for LA with large scale random and uncoordinated deployment
Works with very limited signaling exchange
Needs no additional measurements than what is already established for the LTE
The main features of the algorithm are scalability and operation simplicity
References
H. Murai, M. Edvardsson and E. Dahlman, “LTE-Advanced – The Solution for IMT-Advanced”, ERICSSON,
2008.
[2]
IEEE 802.11 B. P. Crow, I. Widjaja, J.G. Kim, P. T. Sakai, “Wireless Local Area Networks”, IEEE
Communications Magazine, September 1997, pp. 116-126.
[3]
ETSI MCC, “Report of 3GPP TSG RAN IMT-Advanced Workshop”, April 7-8, 2008.
[4]
Arne Simonsson, “Frequency Reuse and Intercell Interference Co-ordination in E-UTRA”, IEEE VTC2007Spring, pp. 3091-3095
[5] R. Giuliano, C. Monti and P. Loreti, “Wireless Technologies Advances for Emergency and Rural
Communications - WiMAX Fractional Frequency Reuse for Rural Environments”, IEEE Wireless
Communications, June 2008, pp. 60-65
[6] T. S. Rappaport and R. A. Brickhouse, “A Simulation Study of Urban In-building Cellular Frequency
Reuse”, IEEE Personal Communications, February 1997, pp. 19-23
[7] IST-4-027756 WINNER II, D1.1.2 “WINNER II Channel Models part I- Channel Models”, Sept 2007.
[8] P. Mogensen, W. Na, I. Kovács, F. Frederiksen, A. Pokhariyal, K. Pedersen, T. Kolding, K. Hugl and M.
Kuusela, “LTE Capacity compared to the Shannon Bound”, IEEE VTC2007-Spring, pp. 1234-1238.
[9] TR 101 112 V3.2.0 (1998-04), Technical Report, Universal Mobile Telecommunications System (UMTS);
Selection Procedures for the Choice of Radio Transmission Technologies of the UMTS (UMTS 30.03
version 3.2.0)
[10] 3GPP TR 25.814 Technical Specification Group Radio Access Network; Physical Layer Aspects for
Evolved UTRA, V7.0.0 (2006-6).
[11] Sanjay Kumar, “Techniques for Efficient Spectrum Usage for Next Generation Mobile Communication
Networks: An LTE and LTE-A case Study” a PhD thesis at Aalborg University, Denmark, June, 2009 (ISBN
978-87-92328-29-8).
[1]
17
Thank You
18