Radio Interface Basics

Wireless transmissions within mobile networks make use of electromagnetic waves to carry the transmitted information from source to destination. The data is modulated to complex waves (in a mathematical sense) using a modulation scheme. Furthermore, a duplex method and a multiple access scheme or multiplex scheme are applied to those complex waves. The resulting modulated spectrum is called baseband. The baseband carries all information and the utilized bandwidth of this baseband depends on the amount of information and spectral efficiency of the modulation scheme. The spectral efficiency is measured in bits per hertz. The baseband is multiplied by a carrier frequency, resulting in a frequency shift in the amount of the carrier frequency. This signal is amplified with an RF amplifier and is transmitted via an antenna. Before the signal is received by the receiving station, the electromagnetic waves carrying the information are distorted by the wireless channel.

The wireless channel is characterized by various time-variant and time-invariant parameters. This section gives only a short introduction to the characteristics of the wireless channel. The properties of the mobile wireless channel can be roughly characterized by the following two attributes:

• Large-scale fading: This is due to loss of signal strength by distance and shadowing of large objects like hills or buildings. It is typically frequency independent, but a function of time and space which fluctuates by means of cell areas.
• Small-scale fading: This is due to the constructive and destructive interference of the multiple signal paths between the transmitting node and the receiving one, resulting in signal strength changes on a spatial scale of the wavelength. Therefore, signal strength variation greatly increases with faster moving stations and is frequency selective.

Free space attenuation as a function of distance d and wavelength λ is shown in the following equation. This basic attenuation (part of large-scale fading) denotes the signal decrease between source and destination without taking any shadowing, multipath fading, or scattering into account:

Especially, the small-scale fading introduces distortion in the received signal to such an extent that it needs to be eliminated, or at least reduced by entities called channel estimator and channel equalizer described later in this section.

Figure 1 shows a typical received wireless channel quality as a function of frequency and time. The channel quality (received signal strength of certain frequencies) changes on a large scale (large-scale fading) and is superposed by the small-scale fading of a moving node. The physical layer of mobile wireless transmission systems have to deal with these characteristics of mobile channels to ensure data transmission to a specific subscriber velocity. Because small-scale fading spatially changes by means of the wavelength, typically by several centimeters in mobile networks, the user velocity introduces fast fading to the received signal. As an example, a user velocity of 100 km/h (27.8 m/s) can result in signal fading changes of 250 times a second. As a result, the received signal is additionally amplitude modulated (a fast amplitude change) caused by the fast fading of a moving user. The amplitude modulated due to the time-variant wireless channel causes additionally frequency dispersion. Entities like fast power control and fast frequency-selective scheduling are introduced into mobile systems in order to counteract this.

 
Figure 1: Time-variant frequency-selective wireless channel. Reproduced with permission from Nomor

A multipath channel is time dispersive, which means that a single transmitted signal is received more than once with different strong echoes (reflections). The electromagnetic waves are reflected by obstacles like buildings, hills, and mountains. The direct beam between the transmit antenna and receiver is called the LOS (Line Of Sight). The LOS is usually the strongest pattern within the brought field of received reflections. h(τ, t) is the time-variant (t and τ) impulse response of the wireless channel with i paths (reflections) and an attenuation a_i(t) of each path i. The impulse response denotes the characteristic behavior of a system (in a mathematical sense), like the wireless channel when an impulse (delta peak) is given as input. In theory, this impulse is so steep that all possible frequencies are included, thus it shows the behavior of all transmitted patterns. τ is the delay of the signal between the source and receiver. This delay is called the propagation delay. τ_i is the additional delay of the reflection path i. Thus,

The transmitted baseband signal is distorted by the wireless channel because different reflections of the signal are interfering at the receiver. Thus, the time-continuous received complex baseband signal yb(t) of a transmitted signal x_b(t) with additive white Gaussian noise n(t) is

f_c is the frequency which is used to transmit the information. It is referred to as the carrier frequency, as mentioned above. The term exp[−j2πf_cτ_i (t)] denotes the time-variant phase shift of each reflection path i.

By knowing the distortion h(τ, t) which was applied to the received signal due to wireless transmission, the receiving entity is able to reverse this distortion and retrieve the transmitted information. In order to estimate the wireless impulse response by the receiver, the transmitter entity inserts known patterns into the transmit signal. Those signals are referred to as pilot or reference signals. Additionally, pilot, reference, or synchronization signals can be applied for time and frame synchronization (in LTE special synchronization signals are used. The receiver scans the received signal for the pilot or synchronization signals by using correlation functions. Once frame synchronization is established, the channel estimator unit of the receiver analyzes the known signal part in order to estimate h(τ, t).

The channel estimator passes the results of the process of estimation to the channel equalizer. The channel equalizer is the entity which removes the distortion due to wireless transmission. Thus, the quality of channel equalization depends on the provided information of the estimator unit. Almost every mobile cell phone standard uses different estimators and equalizers depending on the structure of reference symbols and slot structure. In LTE, channel estimation and equalization are done in the frequency domain and interpolated between adjacent time domain transmission symbols, resulting in a two-dimensional channel equalization.