CRT Sequences With Applications to Collision Channels Allowing Successive Interference Cancellation

Chen, Yi; Lo, Yuan-Hsun; Shum, Kenneth W.; Wong, Wing Shing; Zhang, Yijin

doi:10.1109/tit.2018.2801888

Cited by 16 publications

(9 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Metric Function: For a sequence set C = c m j [n] : j = 0, 1, ..., J −1; m = 0, 1, ..., M −1; n = 0, 1, ..., N−1 consisting of M J sequences of the same length N , the metric function M(C) below 16 reflects how good C is for MC-CDMA systems: (5) Note that our desired metric value M * = inf M(C) = 0. For AlphaSeq, the objective is then to discover the sequence set that minimizes this metric function.…”

Section: Alphaseq For Mc-cdmamentioning

confidence: 99%

“…Remark: In this implementation, the value range of M(s), i.e., [16/9N 3 , 37] ≈ [0, 37], is segmented to three small ranges [0, 15], [5,25], and [10,37]. In the first 8100 episodes, the linear reward is defined in the first small range [0, 15]: metric 0 corresponds to reward −1, and 15 corresponds to reward 1; from episode 8101 to 11400, the linear reward is defined in the second small range [5,25]; after episode 11401, the linear reward is defined in the last small range [10,37].…”

Section: Performance Evaluationmentioning

confidence: 99%

“…This subsection derives the supremum of M(C) in (5) in Section III. Given the definitions of the correlation functions in (2), (3), and (4), we first rewrite (5) as follows: 36 For the second term in (24), we have…”

Section: Appendix Bmentioning

confidence: 99%

“…In this case, the analytical solution is only valid for a subset of sequences of interest. For example, the protocol sequences in [5] are constructed by means of the Chinese Remainder Theorem (CRT) [6]; hence, the number of supported users is restricted to a prime number.…”

Section: Introductionmentioning

confidence: 99%

“…As AlphaSeq plays more and more episodes of the game, we progressively raise the target so that fewer and fewer sequence sets have rewards close to 1, with more having rewards close to −1. In other 5 words, the game becomes more and more demanding as AlphaSeq, starting as a novice, learns to become an expert player.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

AlphaSeq: Sequence Discovery With Deep Reinforcement Learning

Shao

Liew

Wang

2020

IEEE Trans. Neural Netw. Learning Syst.

View full text Add to dashboard Cite

Sequences play an important role in many applications and systems. Discovering sequences with desired properties has long been an interesting intellectual pursuit. This paper puts forth a new paradigm, AlphaSeq, to discover desired sequences algorithmically using deep reinforcement learning (DRL) techniques. AlphaSeq treats the sequence discovery problem as an episodic symbol-filling game, in which a player fills symbols in the vacant positions of a sequence set sequentially during an episode of the game.Each episode ends with a completely-filled sequence set, upon which a reward is given based on the desirability of the sequence set. AlphaSeq models the game as a Markov Decision Process (MDP), and adapts the DRL framework of AlphaGo to solve the MDP. Sequences discovered improve progressively as AlphaSeq, starting as a novice, learns to become an expert game player through many episodes of game playing. Compared with traditional sequence construction by mathematical tools, AlphaSeq is particularly suitable for problems with complex objectives intractable to mathematical analysis.We demonstrate the searching capabilities of AlphaSeq in two applications: 1) AlphaSeq successfully rediscovers a set of ideal complementary codes that can zero-force all potential interferences in multicarrier CDMA systems. 2) AlphaSeq discovers new sequences that triple the signal-to-interference ratio -benchmarked against the well-known Legendre sequence -of a mismatched filter estimator in pulse compression radar systems. Index TermsDeep reinforcement learning, deep neural network, monte-carlo tree search, AlphaGo, multi-carrier CDMA, pulse compression radar.

show abstract

Section: Alphaseq For Mc-cdmamentioning

confidence: 99%