Abstract

Zaporozhchenko Victoriya

Master of Donetsk National Technical University

Faculty: Computer Information Technologies and Automation (CITA)

Department: Automation and Telecommunication (AT)

Speciality: Telecommunication System and Networks (TCS)

theme of Master's Work: Research characteristics of zonal coverage in LTE-based access radio network

Scientific Supervisor: Ph.D., assistant professor of the department AT Vladimir Besarab

ABSTRACT

of the Master's Qualification Work

"Research characteristics of zonal coverage in LTE-based access radio network "

CONTENT

Introduction

1 A theme urgency

2 Communication of work with scientific programs, plans, themes

3 the purpose and problems of the research

4 Prospective scientific novelty of the received results

5 the review of developments and researches on a theme

5.1 At the local level

5.2 At the national level

5.4 At the global level

6. the description of the received and planned results of the work

Conclusions

Bibliography

Note

Introduction

Now the committee 3GPP develops standard LTE for creation of a mobile communication technology which would provide higher speed of data transmission and throughput for "heavy" appendices, such as video on demand, broadband access to the Internet and others. And as in the world the tendency to increase of demand for mobile broadband communication, that is marked, in my opinion, LTE will receive an additional impulse in development.

LTE is a new technology, largely in the state of standardization. this means that it is very difficult to find the references and previous works on this subject. Mostly, 3GPP standardization documents and drafts have to be relied up on though there are articles and books in which to be resulted the review of the given technology.

1 Актуальность темы

For today of a network based on technology LTE in the world had development, there are first practical realizations of small networks, but while experimental samples. In this connection to allocate practical problems of the given technology it is not obviously possible. Characteristics of a covering for radio networks are one of the major indicators of their working capacity. the considered technology is focused on the appendices sensitive to a delay and throughput, therefore a communication quality question in it costs especially sharply. To raise a communication quality allows put in LTE mechanisms of scheduling of a radio channel and adaptation to its conditions. therefore research of characteristics of a covering, development of optimum criterion of adaptation is for today actual, especially in the light of incompleteness of work on standard LTE.

2 Communication of work with scientific programs, plans, themes

Master's qualification work is executed during 2009-2010 agrees with a scientific direction of the department “Computer Engineering” of Donetsk National Technical University.

3 the purpose and problems of the research

3.1 the work purpose

Research objectives: studying radio interface LTE analysis, construction of a radio channel model and research on its example of the basic characteristics zonal coverings, and also working out of scheduling algorithms and development of optimum adaptation criterion and method of transfer to radio channel conditions.

3.2 Idea of work

Idea of work is use of several models of radio signal propagation in space for modelling of radio channel LTE. And also use of the received model for statistics reception on which basis it is possible to develop effective algorithm of scheduling and adaptation of a transfer method to conditions a radio channel.

3.3 the main problems of the research

ДFor realization of idea and purpose achievement master’s works are put and following problems are solved:

3.4. the subject of the research

Subject of research is the mathematical model of a radio channel LTE and the scheduling and adaptation algorithm developed on its based.

3.5. the object of the research

Object of the research is radio wave propagation models and the characteristic of a radio networks covering

3.6. Methodology and methods of researches

In the given work the method of mathematical modelling was applied to radio channel research.

4. Prospective scientific novelty of the received results

Novelty masters works consists in the following: efficiency of an information transfer on radio channel LTE at the expense of development scheduling and adaptation algorithm and selection of optimum adaptation criteria to transfer conditions raises.

5. the review of developments and researches on a theme

5.1. At the local level

LTE technology – new technology, so don’t investigated a lot yet. On our chair researches connected with radio interface LTE before were not conducted, but researches concerning other radio access technologies, radio wave propagation model, researches of radio channel characteristics were conducted:

5.2 At the national level

On on national level concerning technologists LTE have started to be engaged in workings out and researches not so long ago. For today to me following researches in the field are known:

5.3. At the global level

In the world researches and workings out of standard LTE are actively conducted. 3GPP, uniting six organisations from the countries of Asia, the North America and Europe on standardization in the field of working out and advancement of modern mobile technologies. Leaders in this area are Ericsson and Huawei which conduct researches since 2004. Lately, in connection with interest mobile operators in this technology and evolution ways to networks LTE, in territory of Russia as carry out researches, conferences and seminars devoted to the given standard and which are discussed not only practical methods of introduction of these networks, but also theoretical aspects of modelling, radio-frequency planning and an estimation of characteristics of a covering.

6. the description of the received and planned results of the work

My work on this theme has begun with preliminary studying of technology basically on committee materials 3GPP, and also white paper from Nokia, Simens and Ericsson. During such preparation I have drawn for myself a conclusion that a distinctive feature of technology LTE is that it urged not to struggle with problems of a radio channel arising at an information transfer, and to adapt to its changes. To investigate these mechanisms it is necessary to begin with radio wave propagation model in space.

Research methods can be divided, on two groups [6]: numerical considering direct interaction with substance and arising in this connection effects (diffraction, reflexion, refraction, etc.), and semi phenomenological, based on introduction of empirically certain factors of attenuation for this or that type of a landscape with various degree anthropogenic which differ more simplicity of use..

It is necessary to notice that correct application of this or that model allows to receive exact enough results that is important, in a case, for example, there are enough cities, when the direct account of all necessary parameters, such as: height of each building, number of storey’s, width of streets it is etc. almost impossible.

Now researchers allocate three groups of calculation models (methods) of a radio network cover zone [7]:

Models concern statistical methods of calculation the Okamura-Hata, COST 231, Wolfish-Ikegami and so forth in their basis lie the generalized statistical formulas of radio signal attenuation in various types of building (city, suburban, rural). Accuracy of calculation depends on the careful empirical factors based selection on the district maps analysis. At the moment the majority of cellular communication operators uses the software products based on these models, however imperfection of computer district maps for networks planning and average factors lead to enough big error. It is possible to carry rather small time calculation to number of the given models advantages.

the determined methods of cover zones calculation are based on use of physical radio wave propagation model. In them absorption, refraction of electromagnetic waves are considered easing in free space, reflexion from local objects, diffraction on obstacles. Calculation is based on multibeam radio wave propagation model. Plus of this technology is enough split-hair calculation accuracy. In practice the determined method practically is not applied, since in the conditions of city building with difficult architecture covering calculation occupies the large quantity of time comparable in due course of a network expansion. It is caused not only computing expenses for calculation of multibeam radio wave propagation, but also exact modelling necessity of the city environment taking into account architectural features, structures materials that is almost impossible in large settlement scales.

the quasidetermined method differs following features: the multibeam radio wave propagation model is applied, refraction is replaced with easing; average factors of reflexion for each range of frequencies are used; absorption pays off with the account of length of a beam in a structure/wood/park; the adaptive the calculation algorithm considering various direct arrival directions and reflected waves is used; there is a possibility of the each aerial an orientation diagrammes account. the given model has the big accuracy in comparison with statistical methods, however computing expenses much more.

In the documentation on LTE use for modeling of 3 scenarios of radio wave propagation, depending on distances between base stations (BS) and other conditions is provided::

For each of scenarios it is necessary to use the radio wave propagation model in space:

I offer the comparative models characteristic in Table 1 models restrictions, and in Table 2 – a scope are resulted.

Table 1

Model Frequency range, MHz Height of the transferring aerial, m Height of the reception aerial, m Distance between aerials, km
Okumura-Hata 150…1500 30…200 1…10 1…20
COST231-Hata 1500…2000 30…200 1…10 1…20
Walfish-Ikegami 800…2000 4…50 1…3 0.02…5

Table 2

Model Big city Cities of the average or small sizes Residential suburb Countryside
Okumura-Hata + + + +
COST231-Hata + + - +
Walfish-Ikegami + + - -

then for modeling of a radio channel for technology LTE it is possible to use thus the propagation of a signal described above model:

For scenarios of suburban and city macrocell signal propagation model is based on modified model COST 231-Hata:

model COST231-Hata

where - raising height of the base station (BS) aerial, in meters;

- raising height of the mobile station (МS) aerial, in meters;

- carrier frequency, in MHz;

d – distance between BS and MS, in meters;

С – С – constant, for suburban macrocell С=0 дБ, and for city С=3дБ.

For the scenario of city macrocell signal propagation model is based on model Walfish-Ikegami:p>

model Walfish-Ikegami

the model is used at following assumptions:

, height of buildings 12 m, distance between buildings 50 m, width of streets 25 m, .

Based on given mathematical model in the environment of Mathlab there was the obtained data characterizing signal propagation model in space. According to this data a number of a parity a signal/noise values in the channel will be generated. the channel adaptation algorithm will be based on this number of values and to pass from one-beam transfer to multiantenna at deterioration of the given parameter.

Schedulling

Scheduling is the process of dynamically allocating the physical resources among the UEs based on some set of rules, i.e. scheduling algorithm. The link adaptation in this context refers to rate adaptation or MCS selection depending on CQI. In general, link adaptation can also involve transmission power control. Both, scheduling and link adaptation require the CQI (depending on the scheduling algorithm) as input, the link adaptation requires the scheduler output in order to know which users are scheduled and what RBs are allocated to them, and the output of both scheduler and link adaptation, (i.e. the UE Ids of the scheduled users, the resources allocated and the MCS to be used for transmission), are sent to the UEs via PDCCH.

The scheduler may use various algorithms in order to decide which users are to be scheduled and which resources to be allocated to the scheduled users. These techniques may take different aspects into account such as spectral efficiency and fairness. Some of the basic algorithms that are relevant for this thesis work are described below.

  • Maximum sum rate algorithm

    • The maximum sum rate algorithm (MSR) can achieve the maximum sum rate of all users given a total transmit power constraint .

    If the objective is to obtain as much data as possible throught the system, this algorithm is the optimal solution. We can achieve the maximum sum system capacity if the total thought in each subcarrier is maximized. Consequently, the maximum sum capacity optimization problem can be simplified to each subcarrier optimization problem. The maximum capacity is easily obtained when all available power is allocated to the users with the best channel state in that channel. It is a channel aware scheduling rule that allocates the resource to the user experiencing the best channel gains. This is referred to as greedy optimization. The optimal power allocation is processed with the waterfilling algorithm, in which more power is allocated to strong channel and less power allocated to weak channels . Hence the total sum capacity is maximized by summing up the rate on each of the subcarriers.

    With MSR algorithm, the users close to the base station with excellent channels will be allocated most of system resources and have good data rate. The overall system throughput is maximized b the MSR algorithm. However, some users with bad channel condition will not obtain any system resource resulting in user starvation and will be extremely underserved by an MSR-based scheduling procedure.

  • Maximum fairness algorithm.

    • As discussed above, the MSR algorithm may cause unfairness among users. Fairness can be defined in various ways. Two of the most popular criteria are Min-Max fairness and Kelly’s proportional fairness. An alternative method, the maximum fairness algorithm referred as a max-min problem is proposed to allocate resources to users in order that the minimum user’s data rate is maximized. Since the target is to maximize the minimum data rate, the maximum fairness algorithm is referred to as a max-min problem consequently. Due to the concave objective function, the optimal subcarrier and power allocation is much more difficult to determine that in MSR algorithm. It is particularly difficult to find the optimal solution for subcarrier and power allocation problem. Therefore, as discussed in Section 3.4, low-complexity suboptimal algorithm are necessary, in which the subcarrier and power allocation are done separately. The general method is to allocate the equal power to each subcarrier initially. Then iteratively assign the available subcarrier to a low-rate user with the best channel gain on it . After completing the generally suboptimal subcarrier allocation, the power can be allocated optimally. This algorithm concerns system performance in terms of both the fairness achieved and the total throughput. The drawback of this algorithm is that the rate distribution among users is not flexible and the total throughput is largely limited by the users experiencing bad condition, while allocated with most of resource .

  • Proportional rate constrains algorithm

    • In broadband wireless communication system, different users with multiple services require different specific data rates. The proportional rate constrains algorithm (PRC) is proposed to maximize the total throughput with additional constrains of each user’s data rate proportional to set of predefined system parameters {β_k }_(k=1)^k. The proportional data rate constrains can be expressed in Eq 3:

    Formula of proportional rate constrains

    The achieve data rate Rk can be represented as Eq 4:

    Formulao of data rate


    k is number of users;

    L – number of subcarrier;

    hk,l – channel gain for user k in subcarrier l;

    Pk,l – transmit power allocated for user k in subcarrier l;

    σ2- AWGN power spectrum density;

    B – total transmission bandwidth;

    ρk,n – indicates if subcarrier l used by user k.

    The advantage of this algorithm is that the arbitrary data rates can be obtained by changing the values βk. The optimization problem of this algorithm is also very difficult to solve directly because it involves continuous variables, binary variables and the feasible set is not convex Similar to the MSR algorithm to reduce computational complexity, this optimization problem is also divided to the suboptimal problems in which the subcarrier and power allocation are done separately. The near optimal method is derived and outlined in and the low complexity implementation is developed in .

  • Proportional fairness scheduling

    • These three algorithms discussed above attempt to instantaneously achieve the total throughput, maximum fairness, or preset proportional rates for each user. An alternative algorithm could attempt to obtain this objectives overtime. This would provide significant additional flexibility to the scheduling decision. The scheduler could simply wait for the user experiencing bad channel quality to get close to the base station before transmitting. The long-term channel condition is not concerned by the above discussed algorithm. The most popular scheduling algorithm for this type aspect is proportional fairness scheduling policy , which maximizes the long-term user throughput related to the average channel conditions, and satisfies fairness to certain extend. The PF scheduler is developed to make use of multiuser diversity while maintaining comparable long-term throughput for all users.

    Let Tk(t)denote the average throughput for user k up to slot t, and let Rk(t) represent the instantaneous data rate that user k can achieve at time t the parameter tc control the latency o the system. The latency increases with large tc , resulting in higher sum throughput . The latency decrease if tc is small because the average throughput values more quickly, at the expense o some throughput. The selected user is denoted as k* , with the highest Rk(t)/ Tk(t) for transmission by scheduling decision. The long term, this equivalent to select the user with the highest instantaneous rate relative to its mean rate. The average throughput Tk(t) for all users is then updated according to

    Formula of the avarage value of throughput

    The bad channels are unlikely to be selected for each user because scheduler selects the user with the largest instantaneous data rate relative to its average throughput. The fairness can be improved by giving scheduler priority to the consistently underserved users. The PF scheduling algorithm can maximize the long-term user throughput while maintaining to a certain extent . However, this algorithm can’t support real time services such as voice and real time video streaming services.

  • M-LWDF/EXP

    • The modified largest weighted delay first scheduling algorithm is proposed to support not only real time services but also non-real time services. The objective is the maintain each traffic delay smaller than a predefined threshold value with probability . The requirement of delay and thriugput is defined as PT{Wi>τi}≤δi and Ti>tiWi is the head-of-line packet delay for the queue i,


    τi- the maximal allowable delay threshold,

    δi- мthe maximal allowable probability of exeecting τi,

    Ti- present minimum thriugput threshold .

    In each time slot t, a user i ia selected to serve the queue for wichϒi Wi(t)Ri(t) is maximal.

    Ri(t) – is the channel capacity with respect to flow i;

    ϒi – an arbitary constant;

    ϒi=ai/Ri(t), ai=-(log δi)/ τi, and Ri(t) – the average channel rate with respect to flow i.

    By settings an appropriate value to parametr ϒi, the delay requirement can be satisfied. The M-LWDF schedulling algorithm is proopsed to achieve optimal throughput. The key feature of this algorithm is that a scheduling decision depends on both channel and queue states and can meet the delay requiremetns, while it is difficult to find the optimal ϒi value for each traffic tpe .

    The exponential rule is designed with sismslar structure as the M-LWDF scheme to offer both best effort service and delay sensitive for different users The difference is to include different weights for different type of services. The priority function is characterized as:

    The priority function


    where Ui(n) – priority value of the ith user;

    Each Ui(n) is fundamentally evaluated by the EXP rule including the delay of the head-of-line packet.

    Di(n) - the delay of the head-of-line packet in queue for the ith user at the nth slot ;

    Ti(n) - the supportable data rate of the ith user at the nth slot;

    Ti¯ – the average supportable data rate of the ith user at the nth slot;

    ai=-(log δi)/τi, δi - the reqired maximalpacket drop probability of the user i;

    τi – the maximal head-of-kine delay.

    ¯ aD(n)¯ is definited as the following:

    formula 7


    where N – the numbers ofusers in streaming services.

    The throughput is considered most important for best effer service and dela bound is cinsidered as teh most significant criterion for a streaming service. The EXP scheme can satisfy QoS of the delay sensitive services.

    Link adaptation

Link adaptation and scheduling uses channel quality indicator (CQI) as an input to perform resource allocation and MCS selection. As mentioned before, the CQI is derived from SINR measurements made by the receivers (by the UEs in the downlink and by the eNodeBs in the uplink).

However due to previously mentioned sources of inaccuracies such as quantization, delay, long CQI reporting periods and SINR averaging to reduce transmission overheads, it is beneficial to have some kind of CQI adjustment at eNodeB. The simplest way of doing this is to adjust the CQI values by a certain margin, from now on referred to as the Link Adaptation Margin (LAM) as it is defined in the simulation environment. The adjustment can be written as below,

[CQIeff]=[CQI]-[LAM]

The values are denoted as matrices of arbitrary size, where their sizes may depend on the number of users in the cell, number of resource blocks and number of transmission streams in case of MIMO, etc. CQIeff is the effective CQI value that will be passed to the scheduler and link adaptation.

The LAM can be regarded as an amount by which the CQI is backed off before passing to the scheduler and link adaptation. When the LAM is a positive value, CQIeff will be less than the original CQI. Therefore the link adaptation will tend to select a lower data rate, in other words, a more robust MCS (more conservative) than what it would have selected if not for the CQI adjustment. Similarly, when the LAM is negative, the link adaptation will tend to select a high data rate, in other words, a less robust MCS (more aggressive) than what it would have selected without CQI adjustment. Regardless of whether the LAM is positive or negative, a higher LAM is more conservative than a lower LAM and vice versa.