GSM: GSM stands for Global System for Mobile Communication. It is the most successful second generation digital mobile radio system. GSM became popular very quickly because it provided improved speech quality and through a uniform international standard, made it possible to use a single telephone number and mobile unit around the world. The European Telecommunications Standardization Institute (ETSI) adopted the GSM standard in 1991, and GSM is now used in 135 countries.

GSM also stands out from other technologies with its wide range of services i.e.

• Telephony, asynchronous and synchronous data services (2.4/ 4.8/ 9.6 kbit/s), access to packet data network (X.25), telematic services (SMS, fax, videotext etc.), many value added features (call forwarding, caller ID, voice mailbox) and, e-mail and internet connections.

Speech signal processing of GSM is shown below:

Speech coder:

In transmitter section, first of all, data in the form of speech supplied to speech coder where, speech coder convert that raw data or speech firstly into digitized form and then  encoded data using source encoder technique. The speech coder is based on a type of linear predictive coding (LPC) called residual pulse excited linear predictive coding (RPE-LPC). It delivers 260 bits in each 20 msec time interval, hence a bit rateof 13 kbps as shown in fig.

Channel encoder: After speech encoder, coded data or information entered to the channel encoder device where channel encoding takes place so that coded data or information reached their destination safely without added any corrupted data.

Bit interleaving: After channel encoding, data entered into bit interleaving device. Bit interleaving is a technique for making forward error correction more robust with respect to burst errors. We know that the significant bits are encoded by a rate ½, constraint length, L=5 convolutional encoder, and coded and un-coded bits are blocked interleaved to produce data at a rate of 22.8 kbps.

Encryption and Burst assembly: After error correction, the coded bit stream is encrypted in encryption device and then organised for burst transmission in time slots that carry 114 coded bits and some overhead bits.

Modulation: The modulation used to transmit the bits in each time slot is GMSK with BT=0.3 which is a signal pulse. The output of GMSK modulator is translated in frequency to the desired carrier frequency, which is hopped to a different frequency in each frame.

At the receiver end:  the received signal from transmitter end is de-hopped and translated to baseband, thus creating in phase (I) and quadrature (Q) signal components, which are sampled and buffered. The 26 known transmitted bits are used to measure the channel characteristics and, thus, to specify the matched filter to the channel corrupted signal. The data bits are passed through the matched filter, and then output is processed by a channel equalizer, which may be realized either as a decision feedback equalizer or an ML sequence detector that is efficiently implemented via Viterbi algorithm.

The bits in a burst at the output of equalizer are de-assembled, de-encrypted, de-interleaved and passed to the channel decoder. The decoded bits are used to synthesize the speech signal that was encoded via RPE-LPC.

GSM:

GSM stands for Global System for Mobile Communication. It is the most successful second generation digital mobile radio system. GSM became popular very quickly because it provided improved speech quality and through a uniform international standard, made it possible to use a single telephone number and mobile unit around the world. The European Telecommunications Standardization Institute (ETSI) adopted the GSM standard in 1991, and GSM is now used in 135 countries.

GSM also stands out from other technologies with its wide range of services i.e.

• Telephony, asynchronous and synchronous data services (2.4/ 4.8/ 9.6 kbit/s), access to packet data network (X.25), telematic services (SMS, fax, videotext etc.), many value added features (call forwarding, caller ID, voice mailbox) and, e-mail and internet connections.

Speech signal processing of GSM is shown below:

Speech coder:

In transmitter section, first of all, data in the form of speech supplied to speech coder where, speech coder convert that raw data or speech firstly into digitized form and then  encoded data using source encoder technique. The speech coder is based on a type of linear predictive coding (LPC) called residual pulse excited linear predictive coding (RPE-LPC). It delivers 260 bits in each 20 msec time interval, hence a bit rate of 13 kbps as shown in fig.

Channel encoder: After speech encoder, coded data or information entered to the channel encoder device where channel encoding takes place so that coded data or information reached their destination safely without added any corrupted data.

Bit interleaving: After channel encoding, data entered into bit interleaving device. Bit interleaving is a technique for making forward error correction more robust with respect to burst errors. We know that the significant bits are encoded by a rate ½, constraint length, L=5 convolutional encoder, and coded and un-coded bits are blocked interleaved to produce data at a rate of 22.8 kbps.

Encryption and Burst assembly: After error correction, the coded bit stream is encrypted in encryption device and then organised for burst transmission in time slots that carry 114 coded bits and some overhead bits.

Modulation: The modulation used to transmit the bits in each time slot is GMSK with BT=0.3 which is a signal pulse. The output of GMSK modulator is translated in frequency to the desired carrier frequency, which is hopped to a different frequency in each frame.

At the receiver end:  the received signal from transmitter end is de-hopped and translated to baseband, thus creating in phase (I) and quadrature (Q) signal components, which are sampled and buffered. The 26 known transmitted bits are used to measure the channel characteristics and, thus, to specify the matched filter to the channel corrupted signal. The data bits are passed through the matched filter, and then output is processed by a channel equalizer, which may be realized either as a decision feedback equalizer or an ML sequence detector that is efficiently implemented via Viterbi algorithm.

The bits in a burst at the output of equalizer are de-assembled, de-encrypted, de-interleaved and passed to the channel decoder. The decoded bits are used to synthesize the speech signal that was encoded via RPE-LPC.

Signal processing in GSM:

It is done in six steps as given below.

(1)Speech coding(GSM speech coder).

As we know that in GSM we have limited available frequency spectrum, so it is necessary to efficiently use the available spectrum.

So here first preprocessing is performed on analog speech, where the analog speech is first pre emphasized.

The pre emphasized speech is then passed through segmentation block where 160 samples for every 20ms speech is made.T

he samples are then passed through hamming window to reduce the noise interference.

The speech coder used in GSM is RPE-LPC (Residual Pulse Excited-Linear Predictive Coder).

Here first short term prediction is performed on speech for 16 samples.After the short term prediction 8 STP coefficients are derived and coded using 36 bits.Then long term prediction is performed for 20 to 120 samples.

After the long term prediction 8 LTP co efficient are derived which are coded using 36 bits.

Then RPE computation step is performed and RPE coefficient are derived which are coded using 32 bits.

Then only 52 pulses are used which are encoded using 3 bits so total bits are 156 for pulses.

So after complete process of speech coding there will be total 260 bits for 20 ms.

So it is possible to obtain bit rate of 13 kbps.

(2)Channel coding:

Channel coding step is performed to detect and sometimes correct error introduced in the signal.

In channel coding of speech signal extra 196 bits are added so there will be total 456 bits.

So it is also possible to have bit rate of 22.8 kbps after channel coding.

(3)Interleaving:

Here whole 456 bits are spreaded over different time slots of TDMA channel.

The whole 456 bits are arrangedin matrix in the form of 8 column and 57 rows.

Then bits of each column are combined and first column having number(0,8,16...448), the second column having number(1,9,17.....449) and so on.

So after interleaving there will be 8 blocks of 57 bits.

(4)Burst formatting:

In burst formatting step any two block of 57 bits are placed in the one time slot (burst) of the TDMA channel.

(5)Ciphering:

This step adds security to the speech by adding cipher key to the original data.

Here when mobile wants to make a call randomly generated number of 128 bits are transmitted to mobile station.

The mobile station then combines that number with authentication key Ki having also 128 bits in A3 algorithm.

The output signal is called cipher key Kc which is then mixed with the signalin A5 algorithm.

The signal is then transmitted to base station.

(6)Modulation:

In the last step of signal processing the speech signal is modulated by GMSK modulation.

At the receiving side inverse process is done such as demodulation, deciphering, burst deformatting,de interleaving, channel decoding and speech decoding.