WO2011082713A2

WO2011082713A2 - Nonlinearity compensation in optical communication systems via iterative signal clipping and a multiple led transmitter

Info

Publication number: WO2011082713A2
Application number: PCT/EG2010/000001
Authority: WO
Inventors: Hany Abdelmonem Mohamed Elgala; Raed Mesleh
Original assignee: Hany Abdelmonem Mohamed Elgala; Raed Mesleh
Priority date: 2010-01-05
Filing date: 2010-01-05
Publication date: 2011-07-14
Also published as: WO2011082713A3

Abstract

The invention relates to optical indoor wireless, optical outdoor wireless, or optical fiber communication systems, in particular a system utilizing intensity modulated light emitting diode (LED) as a transmitter and a direct detection photodiode (PD) as an optical receiver. The invented iterative signal clipping (ISC) technique is based on iterative clipping of the time domain orthogonal frequency division multiplexing (OFDM) signal and transmission from a multiple LED transmitter. The LEDs are synchronized, located close to each other, and placed to emit light in the same direction. Hence, the channel path gains from each LED to the receiver PD are similar. The received signals from the different LEDs add coherently at the receiver. The effect of distortion due to clipping is eliminated or significantly reduced based on the considered number of LEDs.

Description

NONLINEARITY COMPENSATION IN OPTICAL

COMMUNICATION SYSTEMS VIA ITERATIVE SIGNAL CLIPPING AND A MULTIPLE LED TRANSMITTER

Technical Field:

The invention relates generally to optical communications, and to improvements in the efficiency of optical transmission systems based on intensity modulation (IM) transmitter with direct detection (DD) receiver. Specially, optical communication systems utilizing orthogonal frequency division multiplexing (OFDM), and related techniques, in order to mitigate transmission impairments. In particular, the invention relates to a method to reduce signal clipping distortion by the use of iterative signal clipping (ISC) and multiple light emitting diodes (LEDs) at the transmitter. The LEDs are placed to emit light in the same direction and, therefore, undergoes similar channel conditions.

Background Art:

Orthogonal frequency division multiplexing (OFDM) is a scheme in which digital data is transmitted over large number of orthogonal subcarriers. In particular, the data to be transmitted is encoded into several parallel data streams each of which is assigned to one subcarrier. A conventional modulation scheme, such as quadrature amplitude modulation (QAM) or phase shift keying (PS ) modulates the available subcarriers. OFDM systems are able to support high speed communication links through the use of high order constellations. Attaching a cyclic prefix (CP) to the transmitted OFDM symbols converts the linear convolution of the channel with the OFDM signal to a circular convolution. As a result, the time-varying channel can be easily estimated using frequency-domain channel estimation and a simple frequency domain equalizer can be employed to retrieve the transmitted data. The CP also acts as a guard interval to eliminate inter-symbol interference (ISI) due to multipath dispersion effects. Adaptive modulation per subcarrier can be applied based on the uplink/downlink requested data rates and quality of service (QoS). Multiple access schemes such as time division multiple access (TDMA), frequency division multiple access (FDMA), and code division multiple access (CDMA) can be easily combined with OFDM to provide multiuser access. Furthermore, OFDM systems may be implemented very efficiently by utilizing an inverse fast Fourier transform (IFFT) at the transmitting end to generate the OFDM signal, and a forward FFT at the receiving end to recover the narrowband subcamers. Recently, there has been increasing interest in utilizing OFDM and related techniques within optical communications. However, the deployment of OFDM within optical systems still has some design challenges. Nonlinearity factors have a major impact on system performance. In particular, large signal amplitudes at the transmitter side are clipped if the maximum power reaches the saturation point of the amplifiers, due to the limited dynamic range of the digital-to-analog converters (DACs), or as a result of light emitting diodes (LEDs) nonlinearities. Therefore, the characteristics of these analog devices will have significant impact on system performance, which is very serious in OFDM systems with a large peak-to-average power ratio (PAPR). High peak signal values in OFDM stem from the superposition of a large number of usually statistically independent subcamers that can constructively sum up to high signal peaks in the time domain. As a result, the OFDM signal suffers from significant in-band and out-of-band distortions due to nonlinearities introduced at the transmitter. The in-band component determines the system bit-error ratio (BER) degradation, whereas the out-of-band component affects adjacent frequency bands. On solution is to back-off the average signal power. Alternatively, the reduction of the PAPR through methods such as clipping, filtering, constrained coding, and selective mapping can be considered. However, neither power back-off nor PAPR reduction techniques necessarily result in an improvement in system performance and trade-offs must be considered. Power back-off might result in a significant power efficiency penalty and can significantly compromise signal coverage. PAPR reduction techniques increase the system complexity and/or sacrifice bandwidth efficiency.

Accordingly, it is an object of the present invention to provide a method and apparatus enabling improved efficiency of data transmission in optical OFDM systems through the iterative signal clipping (ISC) technique and a multiple LED transmitter. The transmitter is based on LEDs emitting in the visible, infrared, or ultra violet spectrum. In particular the technique is very beneficial for optical systems combined with OFDM as they extremely suffer from nonlinearity distortion.

SUMMARY OF THE INVENTION:

It is an object of the present invention to provide a method to compensate for the resulting clipping distortion as a result of the light emitting diode (LED) operational constrains.

According to an embodiment of the invention, a method for processing the signal before transmission and transmitting from multiple LEDs of a similar type is provided. The optical communication system iteratively clips the signal before transmission. The signal is clipped at the LED turn-on voltage (TOV) and/or at the maximum permissible voltage which corresponds to the maximum permissible AC current according to the LED manufacturer data sheet. The signal after clipping is transmitted from LED number one. The portion of the signal beyond the dynamic range of the LED is simultaneously transmitted from LED number two. However, If amplitude variations are still beyond the dynamic range of the second LED, the signal is again clipped and LED number three transmits the portion of the signal beyond the dynamic range of the second LED. Clipping is iteratively repeated until all signals to be transmitted from all LEDs are within the LED dynamic range. The number of LEDs is variable and depends on the LED characteristics and the variations of the signal envelope. The LEDs are synchronized to start transmission at the same time. Also, the LEDs are similar, placed to emit light in the same direction and thus exhibits similar channel attenuations and path delays. Optical channel simulation of a transmitter consisting of four LEDs forming a square array with 5cm side length is conducted. Each LED has a beam angle of 30o. The obtained results show that the channel paths are almost identical. Hence, the signals transmitted from the different LEDs add coherently and detection techniques can be applied.

For determining the lower and the higher clipping values, the LED characteristic curve is obtained either through measurements or from the data sheet. The lower clipping value is at the LED TOV. The higher clipping value is at the maximum permissible voltage which corresponds to the maximum permissible AC current. The signal after clipping is transmitted from one LED. The portion of the signal beyond the dynamic range of the LED is processed by the same algorithm and if there are peaks lower than the LED TOV or higher that the maximum permissible voltage, the signal is clipped again and is transmitted from a second LED. The process is repeated until the maximum peak of the signal is less than the maximum permissible voltage and the lower peak of the signal is higher than the TOV. All clipped signals are transmitted at the same time form the different LEDs.

These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter.

Disclosure Of Invention:

Figure 1 shows a system model of an optical communication system based on the invented iterative signal clipping (ISC) technique and a multiple light emitting diode (LED) transmitter. The incoming data bits to be transmitted by one OFDM symbol/frame from the source bits 30 are modulated using a quadrature amplitude modulation (QAM), phase shift keying (PSK), or pulse

x (k ) amplitude modulation (PAM) modulator 31 resulting in the symbol vector ' . The symbols are assigned to subcarriers 32 to produce real signal at the output of the inverse fast Fourier transform (IFFT) (OFDM modulator) 34 as follows:

DC-biased OFDM:

s^{] (k) = [0 x(k) 0 x^* (N - k)]^~ £ = l : (N/2) -l

ACO-OFDM:

s^{2} (k) = [0 x(k) 0

£ = l : 2 : (N/2)-l

A real bipolar OFDM signal can be generated by constraining the input to the IFFT operation/ OFDM modulator 34 to have Hermitian symmetry. The DC

( ⁵ (^) ) and the ^S(^{N / 2}) subcarriers are set to zero to ensure that ^s^ is symmetrical. For DC-biased OFDM systems with ^N subcarriers as in equation

1 , the complex symbols *W are mapped to subcarriers, ^ = ^{1 :} (^/²) ^{_ 1} . For ACO-OFDM systems, the same procedure is applied except that ^ ' are only assigned to odd subcarrier, ^{k = 1 : 2 :} (^ ²) ^{_ 1} . The even subcarriers are not used for data transmission and are set to zero. The resulting signal at the output of the IFFT operation/ OFDM modulator 34 is anti-symmetrical which allows clipping at the zero level and transmitting only the positive signal. The invented

s(t)

ISC technique is then applied on ' .

The time domain optical OFDM signal s(t) ' is used to modulate the optical carrier intensity. For DCO-OFDM, the time domain signal is bipolar. Therefore, the LED should be biased before applying the modulating signal since the optical intensity cannot be negative (to create a unipolar signal from the bipolar signal). Alternatively, for ACO-OFDM, the bipolar signal is converted to unipolar through clipping of all negative values at the zero level. The flowchart of the ISC technique is shown in Figure 2.

As a result of the LED characteristics, the lower signal peaks are forcibly clipped at the LED turn-on voltage (TOV) ^v . Also, the upper peaks are purposely clipped before modulating the LED at the maximum permissible AC voltage ^v+ . This is to avoid LED chip overheating, which degrades the output light and may lead to total failure.

For DC-biased OFDM, the samples that are larger than ^v+ are clipped and the samples that are lower than ^v are also clipped. For ACO-OFDM, the samples that are larger than ^v+ are clipped. The aim of the invented ISC technique is to avoid signal distortion caused by clipping. This distortion can be clearly noticed when considering the simulation results depicted in Figures 6 and 8 for ACO-OFDM and DC-biased OFDM systems, respectively. The use of high constellation and large OFDM signal power (to achieve large signal-to-noise-ratio (SNR)) is not possible. Both systems suffer greatly for OFDM average electrical signal powers larger than 14dBm. Therefore, the OFDM signal power has to be backed-off before transmitting through the LED to avoid distortion and to guarantee acceptable performance.

The invented ISC technique avoids power back-off through multiple LEDs and iteratively clipping the signal before transmission (LED intensity modulation).

If ^s^ suffers from clipping, i.e. it is clipped at ^ and ^v . The clipped signal is transmitted from LED 1 (for example, see Figures 4 and 5). The portion of the signal beyond the dynamic range of the LED is processed again by the same algorithm and if it still suffers from clipping, it is again clipped and transmitted from LED 2. The process is repeated until the signal amplitudes is within the dynamic range of the LED ( ^v+ and ^v ) or the maximum number of LEDs is reached. The signals are transmitted simultaneously from all LEDs as follows, g (t) = h_l (t) ® y_l (t) + h₂ (t) ® y₂ (t) +L + h_n {t) ® y_n (t) + w(t) ₍₂₎

where contains the channel path gains from LED ¹ to the received photo detector (PD), ^' ^ is the transmitted signal from LED * , and ^w^ is a noise component representing the thermal noise and the shot noise which can be modelled as an additive white Gaussian noise (AWGN). In equation 2, if the LEDs are close to each other and placed to emit light in the same direction, i.e. same azimuth and elevation angles, the attenuation of the channel paths are very similar. A simulated scenario is depicted in Figure 3. Four identical LEDs are located at the ceiling of a room of height 4m and forming a square-shaped array with 5cm side length. The receiver is located at lm desk height. The transmitters has 30o beam angle. The resulting channel paths are almost identical. Hence, equation 2 can be re-written as follows, g (t) = h (t) ® [_y] (t) + y₂ (t) +L + y„(t)] + w (t) ₍₃₎

where L-^ w j y_n \ j _an(j _me dipping effect is very much reduced. The enhancement can be clearly seen from Figures 7 and 9 for the ISC based ACO-OFDM and the DC-biased OFDM systems, respectively. Only with four LEDs, the OFDM average electrical signal power increases by almost lOdBm. This leads to an increase of SNR and to a wider coverage. The performance can be further enhanced by considering more LEDs. In fact, if the number of the available LEDs is large enough to include all the signal after applying the ISC technique, the performance will be irrelevant to the OFDM average electrical signal power. This means that very large SNR can be achieved even for very high constellations resulting in a wider coverage and optimum performance.

BRIEF DESCRIPTION OF THE DRAWINGS:

Preferred embodiments of the invention are described with reference to the accompanying drawings, in which: Figure 1 illustrates schematically a system for optical orthogonal frequency division multiplexing (OFDM) communication applying the invented iterative signal clipping (ISC) technique with a multiple light emitting diode (LED) transmitter. The system considers any modulation technique (quadrature amplitude modulation (QAM)/ phase shift keying (PSK)/ pulse amplitude modulation (PAM)/ etc ..) and any optical OFDM system (DC-biased OFDM system/ asymmetrical clipped OFDM system (ACO)/ pulse amplitude modulation-discrete multi tone (PAM-DMT)) as determined by the subcarrier assignments.

Figure 2 is a flowchart illustrating the invented ISC technique.

Figure 3 shows the simulated optical channel for simple off-the-shelf LEDs. The transmit unit consists of four LEDs forming a square-shaped array with 5cm side length. The channel matrix and it's histogram plot are shown.

Figure 4 shows example time domain signals of a 256-QAM ACO-OFDM system utilizing the invented ISC technique. The original signal before clipping and the signals to be transmitted from three LEDs are shown.

Figure 5 shows example time domain signals of a 256-QAM DC-biased OFDM utilizing the invented ISC technique. The original signal before clipping and the signals to be transmitted from three LEDs are shown.

Figure 6 shows the bit-error performance of ACO-OFDM suffering from signal clipping. The performance is plotted versus the OFDM signal power for different QAM constellations. Figure 7 shows the bit-error performance (BER) of ACO-OFDM utilizing the invented ISC technique. The performance is plotted versus the OFDM signal power for different QAM constellations.

Figure 8 shows the bit-error performance of DC-biased OFDM suffering from signal clipping. The performance is plotted versus the OFDM signal power for different QAM constellations.

Figure 9 shows the BER of DC-biased OFDM utilizing the invented ISC technique. The performance is plotted versus the OFDM signal power for different QAM constellations.

Claims

1 - Optical communication system comprising an optical transmitter having an electrical input and an optical output. The electrical information-bearing signal modulates a light source and emits optical output wherein there is a nonlinear relationship between input electrical signal amplitude and corresponding output optical amplitude caused by the light emitting diode (LED), the laser diode (LD), the amplifier, the digital-to-analog converter (DAC), or any other analog circuit device or component at the transmitter side.

2- Optical communication system according to claim 1 wherein the information- bearing electrical signal is composed using the optical othogonal frequency division multiplexing (OFDM) or the optical discrete multi tone (DMT) technique.

3- Optical communication system according to any one of claims 1 to 2 wherein the light source is an LED or a LD and, wherein multiple LEDs or LDs are of the same type.

4- Optical communication system according to any one of claims 1 to 3 wherein the LED or the LD is DC-biased prior to transmission of a bipolar signal and, wherein a unipolar signal is transmitted at the LED turn on voltage (TOV) or at the LD threshold voltage and, wherein the signal is iteratively clipped before transmission from multiple LEDs or LDs and, wherein the LEDs or LDs are located close to each other and placed to emit light in the same direction and thus exhibit similar channel conditions. 5- Optical communication system according to any one of claims 1 to 4 wherein the light is non-coherent or coherent.

6- Optical communication system according to any one of claims 1 to 5 wherein the light is infrared light, visible light, or ultraviolet light.

7- Optical communication system according to any one of claims 1 to 6 wherein the optical channel is indoor free-space, outdoor free-space, a single mode optical fiber or a multimode optical fiber.

8- Optical communication system according to any one of claims 1 to 7 wherein the LEDs or LDs are capable of amplitude-modulating by using a digital modulation technique.

9- Optical communication system according to claim 8 wherein the digital modulation technique is selected from the group consisting of quadrature amplitude modulation (QAM), phase shift keying (PSK), and pulse amplitude modulation (PAM).