The sixth generation (6G) communication era will achieve the full coverage of mobile communication networks, covering air, ground and underwater space. In view of the limited spectrum resources of radio frequency, visible light communication (VLC), as a complementary wireless communication technology with radio frequency communication, is expected to become an indispensable part of 6G system. At present, both indoor VLC and underwater VLC have been extensively studied. Light-emitting diode (LED) has become the most commonly used light source in indoor VLC because of its environmental protection and durability, and dual functions of lighting and communication. However, the bandwidth of leds is relatively narrow, so how to improve spectral efficiency has become the key to improving indoor VLC performance. Although VLC does not have the problem of Doppler shift, it still faces the challenge of multipath fading effect caused by different transmission paths. At present, many VLC systems adopt orthogonal frequency division multiplexing (OFDM) technology to deal with multipath effects and intersymbol interference. However, OFDM technology requires the insertion of a loop prefix (CP) before each symbol, which not only increases the length of the 'frame', but also reduces the amount of user data available within each frame, ultimately leading to a decrease in the total capacity of the VLC system.
Orthogonal time-frequency space
Orthogonal time-frequency space (OTFS) modulation can enhance the capacity and reliability of multipath propagation by taking advantage of the orthogonal characteristics of time and frequency. Studies have shown that OTFS has better receiving sensitivity than OFDM modulation at the same spectral efficiency, and OTFS does not need to add a cyclic prefix (CP) before each symbol, which means that the application of OTFS modulation in VLC systems can bring better performance than OFDM. Using the advantages of OTFS modulation, the spectral efficiency and transmission speed of indoor VLC can be effectively improved. However, it has been pointed out that when the number of cyclic prefixes (CP) is insufficient, OTFS will cause serious intersymbol interference when processing visible multipath propagation, which will significantly affect the communication quality of VLC.
VLC channel estimation and equalization technology under OTFS modulation
In order to further improve the transmission rate and communication quality of indoor VLC, VLC channel estimation and equalization technology under OTFS modulation is discussed to reduce the negative impact of intersymbol interference. Least squares (LS) estimation and least mean square error (MMSE) estimation are two commonly used methods for OTFS channel estimation. Among them, the accuracy of LS estimation is low, which can not meet the demand of indoor VLC channel estimation. Although MMSE estimation has high accuracy, it is costly to calculate because it involves a large number of complex matrix inversion operations, and requires some channel prior information, such as channel noise variance, which is difficult to obtain accurately in indoor VLC environment. In recent years, some researches have proposed OTFS channel estimation algorithms based on transform domain basis function and learn-based OTFS channel estimation algorithms. These two methods have made progress in improving the accuracy of OTFS channel estimation and reducing the computational complexity respectively, but they usually need to add additional pilot signals in OTFS system. This not only requires the design of a special frame structure, but also leads to a decrease in the data transmission rate of indoor VLC.
OTFS modulation is used to improve the rate and reliability of indoor VLC, and ResCNN is proposed to accurately estimate and equalize VLC multi-carrier channel, which effectively compensates the influence of intersymbol interference on VLC performance. When the signal transmission distance is 1 m, and the transmission rate is 512 Mb/s, 768 Mb/s, 1 Gb/s, and 1.5 Gb/s, respectively, the bit error rate estimated by ResCNN is lower than 3.8×10−3. The channel estimation method in this paper does not need to increase the cyclic prefix coding and pilot frequency. The residual convolutional neural network is trained through training sequences in the offline stage, and only a small number of training sequences are required to complete the channel estimation in the online stage. Therefore, the system capacity of VLC is guaranteed by this method. However, the proposed method is only applicable to static scenarios and static channels. In the future, the channel estimation and equalization of OTFS indoor VLC in mobile scenarios will be studied, and the application range of the proposed method will be extended to make it play a greater value in 6G mobile communication.