Experimental Demonstration of Bandwidth-Efficient and Low-Complexity Mobile Fronthaul Transmissions Utilizing Digital Orthogonal Filtering-Enabled Channel Aggregation

Abstract

We experimentally demonstrate bandwidth-efficient and low-complexity mobile fronthaul transmissions, where 24 20MHz LTE signals are aggregated via digital orthogonal filtering, achieving an average EVM of 2.8% and 31dB loss budget for a 6km SSMF link.

Rectangular Orthogonal Digital Filter Banks Based on Extended Gaussian Functions

“…The FFT-based frequency domain overlap-add method and the conventional convolution-based time domain method can be used for realizing the digital filtering operation. However, for transmission systems [10]- [12] similar to the PONs considered in this paper, before the digital filtering operation, a digital M× up-sampling operation should be performed via inserting (M-1) zeros between every two consecutive data samples. To support a large number of channel count, a large digital up-sampling factor of M is thus required.…”

“…For the fixed networks of the 5 th generation and beyond [7], to effectively accommodate a large diversity of new use cases with diversified requirements on bandwidth, latency, coverage, QoS and security, it is envisaged that digital filtering acts as a promising physical layer technology to provide the fixed networks with sufficient flexibility, elasticity and adaptability. On the other hand, for the rollout of 5G and beyond mobile networks, advanced digital filtering techniques are also highly desirable for not only realizing the adaptive spectral confinement of waveforms transmitted over the radio/fiber links [8], [9], but also enabling the fronthaul mobile networks to provide, in a cost effective manner, dynamically reconfigurable, on-demand 'just-the-right-size' elastic connections each tailored to a specific application and/or service requirement [10], [11]. Such operations are vital for achieving the seamless convergence of separately developed and operated optical networks and mobile networks [12], [13].…”

“…In all previously published work [5], [6], [10]- [22], [25], [26], to the best knowledge of the authors, the rectangular ODFBs are produced using a prototype filter having a square-root-raised-cosine (SRRC) profile only. To minimise signal distortions due to imperfect digital filter implementation, the SRRC-based ODFBs have to use a sufficiently long digital filter length to achieve desirable digital filter characteristics, i.e., flat passbands, sharp transition bands and sufficiently large stopband attenuations.…”

Analyzing Peak-to-Average Power Ratio Characteristics in Multi-Channel Intensity Modulation and Direct Detection Flexible Transceivers Deploying Inverse Fast Fourier Transform/Fast Fourier Transform-Based Processing

“…For existing DSP-based multi-channel aggregation and de-aggregation techniques, multi-channel aggregation/de-aggregation can be realized with a single inverse fast Fourier transform/fast Fourier transform (IFFT/FFT) operation [ 2 , 3 , 4 ], a digital filtering operation [ 5 , 6 , 7 , 8 ], and code-division multiplexing [ 9 , 10 ]. In comparison with these techniques, the cascaded IFFT/FFT-based multi-channel aggregation/de-aggregation technique [ 11 ] can not only potentially operate at an ‘add-as-you-grow’ mode to support adaptive and flexible variations in both channel count and channel line rate but also offer additional physical layer network security and guard band-free and highly spectrally efficient transmissions.…”