By:

Dr. Ibrahim Fayed

Dr. Eman S. El-Din

Contents

• Introduction to Machine to Machine

• M2M Applications & Network Architecture

• Problem Definition

• Model Flow Chart

• Model Parameters & Assumptions

• Mathematical Analysis

• Results

• Future Work

NTI 2

Introduction to Machine toMachine

-Machine-to-Machine(M2M) involves communicationwithout (or only limited)human intervention

-The human is not the input, but only (optionally) the output

-Also named Machine Type Communication (MTC)in 3GPP

NTI 3

M2M Applications

NTI 4

M2M Network Architecture-scenario1 With MTC Server

Toward Ubiquitous Massive Accesses in 3GPPMachine-to-Machine Communications IEEE CommunicationsMagazine • April 2011

NTI 5

M2M Network architecture-scenario2 without MTC Server

Toward Ubiquitous Massive Accesses in 3GPPMachine-to-Machine Communications IEEE CommunicationsMagazine • April 2011

NTI 6

Problem Definition:

• Overload in access channel due tocongestion in accessing the network(RACH procedure)

• A collision happens if two or more UEstake the same preamble in the samesubframe.

NTI 7

Delta 1

Delta 2

RACH PROCEDURE SUCCESS

NTI 8

Message 1 Congestion

NTI 9

Different Preamble

NTI 10

Delta1

Delta2

Message 3 Failure

NTI 11

Delta1

Delta2

Msg4 buffering

Msg4UE4

Msg4UE2

Msg4UE1

Msg4 Queuing

NTI 12

Model Flow Chart

NTI 13

State Transitioning Diagram

NTI 14

Total Probability of Failure• Total Probability of Failure

•P : probability of collision in msg1 (same preamble transmittedtwice simultaneously)

•g: probability of collision in msg3/4

•n:number of states to retransmit msg 1

•m: number of states to retransmit msg 3

NTI 15

Parameters .

From “Analysis of PDCCH Performance for M2M Traffic in LTE”by Prajwal Osti, Pasi Lassila, Samuli Aalto, Anna Larmo, and Tuomas Tirronen

November 6, 2014.

Symbol parameter Typical Values

K Number of preambles 54

b RACH Periodicity 5

c Max number of UL grants persubframe

3

NTI 16

Assumptions :•To simplify the analysis, we assume that no othertraffic, except this random access traffic, is present inthe network

•All random access requests together (not only the freshones but also the retransmissions) constitute a Poissonprocess, the rate of which is denoted a (attempts persubframe).

•Distribution of users in network follows Uniformdistribution

From “Analysis of PDCCH Performance for M2M Traffic in LTE”by Prajwal Osti, Pasi Lassila, Samuli Aalto, Anna Larmo, and Tuomas Tirronen November 6, 2014.

NTI 17

MATHIMATICAL MODEL

• Pi(a) = Pr{Ank=i}=ೌ್

ೖ

!e-ab/K

• ≔ }ܚ۾

= , ,

~()=j}=

ି ି

− −

ି ≥ ≥ ≥

•

:=Pr

=i, ≈()

=j}

=

ୀୀ

షష

From “Analysis of PDCCH Performance for M2M Traffic in LTE”by Prajwal Osti, Pasi Lassila, Samuli Aalto, Anna Larmo, and Tuomas Tirronen November 6, 2014.

NTI 18

CONT.

• =i}

• భ

• మ

NTI 19

Results and analysisThroughput of Msg 1

Our results

NTI 20

aଵ

Throughput of Msg 2Our results

NTI 21

ܽଶ

Throughput of Msg 1& Msg 2Our results

NTI 22

∗

Verificationb=5, c=3,M=4,M=3

Our results “Analysis of PDCCH Performance for M2M Traffic in LTE”by Prajwal Osti, Pasi Lassila, Samuli Aalto, Anna Larmo, and Tuomas Tirronen November 6, 2014NTI

Future work• Traffic pattern analysis based on bandwidth consumption (ex:

queuing for PDCCH & its impacts on throughput )

• Multi hop connections between eNBs unlike single hop H2Hcommunication.

• Security to maintain data integrity by protecting from potentialhackers

• Promising solution for the air interface of general M2Mcommunications may lie in the cognitive radio technology to reuselicensed spectrum

NTI 24

NTI 25