LTE is the next generation of mobile broadband technology which is based on Multiple Input Multiple Output (MIMO) and Orthogonal Frequency Division Multiplexing (OFDM). It mainly concentrates to satisfy carrier needs for high speed data and media transport as well as high capacity voice support. It provides data rates up to 50Mbps for uplink and 100 Mbps for downlink. It improves capacity and coverage and utilizes bandwidth up to 20MHz of spectrum. When the information is send over a wireless channel, signal gets corrupted due to multipath effect. In order to find the effect of channel on the transmitted signal channel estimation is performed. Channel estimation is a key part of OFDM receiver, since it provides channel information for channel equalization in the receiver side and deciding transmission parameters in the transmitter side. A special reference signal embedded in the Physical Resource Block (PRB) is used for facilitating channel estimation in LTE systems. This paper focuses on the reference signal generation and channel estimation in LTE downlink
Keywords: Reference signal, LTE, Channel estimation, Pilot, Least Square (LS) algorithm.
The needs of human beings in communication fields are increasing day by day. In order to meet these needs 3GPP evolved a new technique called LTE. LTE is incorporated in wireless interface and wireless network. The aim of this LTE standard is to improve speed as well as capacity of wireless data networks using OFDM and MIMO techniques which are new to cellular communication system. Their other design target was to change the network architecture to an IP-based system so that it could reduce the transfer latency as compared to the previous existing standards.
Channel estimation is a significant part for the design of receivers in wireless communication systems. For receiving the transmitted information accurately, it is necessary to know about channel behaviour. So that we should transmit pilot carriers for estimating the channel behaviour. Channel estimation provides the receiver to approximate channel impulse response and thereby behaviour of the channel. In this project ,the pilot based channel estimation algorithm for 3GPP LTE downlink is considered.
The estimation of the channel characteristics in LTE is carried out using cell specific reference signals (pilot symbols) inserted together in time and frequency domain. These pilot symbols give an estimate of the channel at given locations inside a sub frame. Through interpolation it is capable for estimating the channel across an arbitrary number of sub frames.
LTE is the next leap in cellular communication systems. LTE is a 3GPP standard that provides an uplink speed of up to 50Mbps and downlink speed of up to 100Mbps. It is providing a scalable bandwidth from 1.25 MHz to 20 MHz This suits needs of different service providers having different bandwidth allocations. It also helps the operators to provide a variety of functions based on the spectrum available.
LTE makes use of Orthogonal Frequency Division Multiple Access (OFDMA) for down link and Single Carrier Frequency Division Multiple Access (SCFDMA) for uplink. LTE makes use of multiple-antenna techniques such as MIMO which can either raise channel capacity or improve signal robustness in addition to OFDM.
Both OFDM and MIMO are two most important techniques used in LTE and cause major differentiation over 3G systems which are based on code division multiple access (CDMA). The OFDM technology is based on using several narrow band sub-carriers which are mutually orthogonal in the frequency domain which helps to remove Inter Symbol Interference (ISI).However OFDM cause high Peak To Average Ratio (PAPR) hence LTE uses SCFDMA for uplink. This reduces PAPR as there is only a single carrier as opposed to N carriers.
III. Channel estimation
The signal transmitted over the communication channel gets affected by different types of noise and hence signal that we get at the receiver is corrupted .When information is transmitted over a wireless channel; the signal get distorted due to multipath effect. Typically there is a line-of-sight path existing between the transmitter and receiver. In addition, there exist many other paths produced by different obstacles in the path such as signals get reflected by buildings, vehicles and other obstructions. Signals travelling over different paths reach the receiver, but at different time. Therefore there exist a time delay between these signals depending on the distance travelled along each path. Also the channel may cause frequency selective fading.
The signal received is corrupted, so in order to regenerate the original signal we need to find the effect of channel on the signal transmitted. Channel Estimation (CE) technique allows the receiver to approximate the channel impulse response and to find the channel behaviour. Once the channel behaviour is known we can easily regenerate the signal at receiver. Channel estimation can be used to improve signal to noise ratio, for detecting signal at receiver, channel equalization etc.
Channel estimation techniques can be divided into two main categories. Trained channel estimation and Blind channel estimation. In trained channel estimation it makes use of a trained sequence, which is known to the receiver. Blind estimation does not need any trained sequence but it makes use of the statistical information. Comparing to trained channel estimation Blind estimation need less bandwidth but computation is complex, so blind estimation is mainly used in the application of real time systems.
III. Reference signal generation
Cell Specific Reference Signals
In all downlink sub frames which support PDSCH transmission cell-specific reference signals shall be transmitted. It is transmitted on one to four antenna ports (0 to 3). Cell-specific reference signals bandwidth is defined only for 15 kHz.
Figure: 1 Resource block and resource elements (Normal CP)
LTE Evolution Lab has been proposed to help the mapping of physical channel and signals in the resource grid. How a RS is generated and mapped in an empty grid for a single antenna is shown in figure I. For a coherent demodulation pilot symbols are inserted in the OFDM time-frequency grid for the purpose of channel estimation. Downlink RS are inserted within first and third last OFDM symbol of each slot.
In the LTE air interface, Physical layer cell identity is used for cell identification and channel synchronization. The primary and secondary synchronization sequences were determined by the Physical cell ID. It is similar to scrambling codes from UMTS. Physical cell ID has a range from 0 to 503 values; it is used to scramble the data. It is calculated by the equation (1), (2) and (3).
Where, ns -slot number within a radio frame,
l-OFDM symbol number within a slot
Where defines is the group, of cell belongs to (1-167) and defines the identity range (0-2).
We make use of gold sequence for generating the reference signal. First pn sequence is given by (4)
1st pn sequence is initialized with (1,0,0....)
Second pn sequence is given by
2nd pn sequence is initialised using the equation given below.
Figure:2 PN scrambling code generation in LTE system
The gold sequence is given by
The gold sequence generated using equation (7) is used for generating refrence signal using equation (8).
LTE uses cell specific reference signals (pilot symbols) inserted in both time and frequency, for the estimation of the channel characteristics. These pilot symbols will give an estimate of the channel at specified locations inside a sub frame. To estimate the channel across an arbitrary number of sub frames we can use interpolation.
In each physical antenna port cell-specific RS is transmitted. It is used for both demodulation and measurement point. Its pattern design ensures channel estimation correctness.
IV. Least Square (LS) Channel estimation
Reference signal embedded in resource grid is employed for channel estimation. There are several algorithms for channel estimation, among that we use LS algorithm for channel estimation. LS is one of the simplest algorithms for channel estimation.
The goal of channel estimation is to find channel frequency response H by observing the received signal Y, with the assumption that the transmitted signal X is already known (reference signals). We can represent the received signal as
Where W(k) is the Fourier transform of the noise W. So from the above equation (9) we have the LS estimate of H as given below
So by making use of LS algorithm it is possible to calculate channel estimation at the position of reference symbols. Channel estimation on the remaining positions is achieved by using interpolation. Linear interpolation is the easiest interpolation method.
V. Simulation and results
We are considering channel bandwidth of 1.25MHz, with FFT size of 128 and the number of occupied sub carriers is 76. For this we need 6 resource blocks. For the first pn sequence, the initial sequence is fixed as (1, 0, 0...) and for second sequence it varies as given in equation (6).
Reference signal generated is transmitted through an AWGN channel and at the receiver the channel is estimated by making use of the received signal.
First pn sequence
c1 = [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
2nd pn sequence for NID=0, ns=0 ,l=0 and therefore cinit=8193
C2 = [0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
Gold sequence for NID=0,
[0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
Last pn sequence for NID=503, ns=19, l=4 and therefore cinit=149520367
[ 0 0 0 1 1 1 0 1 0 0 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 0 0 0 0]
Gold sequence for Nid=503
[0 0 0 1 1 1 0 1 0 0 1 0 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 0 0 0 0]
re1 = [0.7071 + 0.7071i 0.7071 + 0.7071i 0.7071 + 0.7071i 0.7071 + 0.7071i 0.7071 + 0.7071i 0.7071 - 0.7071i 0.7071 + 0.7071i 0.7071 + 0.7071i 0.7071 + 0.7071i 0.7071 + 0.7071i 0.7071 + 0.7071i 0.7071 + 0.7071i]
Figure 3: Magnitude Response of actual and estimated channel
LTE is the next generation mobile broad band technology with a scalable bandwidth. OFDMA with MIMO technology allows a high speed of 100 Mbps in downlink, while SC-FDMA on the uplink reduces the PAPR values to a certain limit and at the same time it provides uplink speed of 50Mbps. The physical layer (PHY) is well designed to offer high cell-edge performance with specific characteristics such as dynamic bandwidth allocation to users. The design of reference signals take into account the path loss and interference environment at the cell edge. Channel estimation is based on the known training bits (piolet symbols) and corresponding received samples at the receiver. For our channel estimation we use LS algorithm since it is simple and effective for high SNR values.
 Jim Zyren, "Overview of the 3GPP Long Term Evolution Physical Layer," White paper from
Free Scale Semiconductor, July 2007
 'LTE in a nutshell: The Physical Layer', white paper from Telesystem Innovations,2010
 '3GPP Long Term Evolution -Physical Layer'-Prof.S.J. Thiruvengadam, March 2012
 'Tutorial on functionality and performance of air interface of LTE system' -Prashant Srivasthava, University of Texas, December 2010
 Farooq khan, "LTE for 4G mobile broadband Air interface and Performance", Cambridge
University Press, 2009