Pilot scheduling based on water-filling algorithm in massive MIMO

Zhou, Zhichao; Wang, Dong

doi:10.1109/iceiec.2016.7589694

Cited by 9 publications

(6 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…entries with zero mean and variance σ 2 N . 35 The received user data at the n th BS antenna are given by…”

Section: System Modelmentioning

confidence: 99%

“…Using the matched filter (MF) detector, the received symbol c x up jk can be obtained from the channel estimate d h jkn as given below: 35 c…”

Section: System Modelmentioning

confidence: 99%

“…In this case, we have considered N t ≥ K , and N t is limited. Here, it is considered that the K users are using the same time‐frequency resources, and they are uniformly distributed over the entire cell, after leaving a central disc of radius r d around the BS 35 . The j th antenna of the k th user is represented as < j , k >.…”

Section: System Modelmentioning

confidence: 99%

“…Then, the pilot sequence received at the n th antenna of the BS is given as

Y_{n}^{q} = \sqrt{ρ_{q}} true \sum_{\underset{j \in [1, N_{r}]}{k = 1}}^{K} h_{j k n} φ_{p} + N_{n}^{q},

where ρ q denotes the pilot transmitting power and

N_{n}^{q}

represents the Gaussian noise having independent and identically distributed (i.i.d.) entries with zero mean and variance

σ_{N}^{2}

35 . The received user data at the n th BS antenna are given by

y_{n}^{u p} = \sqrt{ρ_{u p}} true \sum_{\underset{j \in [1, N_{r}]}{k = 1}}^{K} h_{j k n} x_{j k}^{u p} + η_{n}^{u p},

where

x_{j k}^{u p}

represents the uplink data transmitted from j th antenna of k th user in the cell and ρ up represents the uplink data transmitting power.…”

Section: System Modelmentioning

confidence: 99%

“…Using the matched filter (MF) detector, the received symbol

\overset{x_{j k}^{u p}}{true}

can be obtained from the channel estimate

\overset{h_{j k n}}{true}

as given below: 35

\begin{matrix} right & left \hat{x_{j k}^{u p}} = \hat{h_{j k n}^{H}} y_{n}^{u p} & right \\ right & left = {(\sum_{k \in Ψ_{k}} h_{j k n} + w_{k n})}^{H} (\sqrt{ρ_{u p}} {falsefalse}_{\sum}^{\underset{k^{'} = 1}{true}} h_{j k^{'} n} x_{j k^{'}}^{u p} + η_{n}^{u p}) \\ right & left = \sqrt{ρ_{u p}} (h_{j k n}^{H} h_{j k n} x_{j k}^{u p} + \sum_{k^{'} \neq k} h_{j k^{'} n}^{H} h_{j k^{'} n} x_{j k^{'}}^{u p}) + ξ_{k n}^{u p}, \end{matrix}

where

ξ_{k n}^{u p ...}

…”

Section: System Modelmentioning

confidence: 99%

See 4 more Smart Citations

Efficient pilot reuse algorithms for massive MIMO cellular system

Dey

Pattanayak

Gurjar

et al. 2020

Int J Communication

View full text Add to dashboard Cite

Summary Pilot contamination has come up as one of the major limitations of massive multiple‐input multiple‐output. In the recent past, the pilot contamination resulted by the pilot reuse in inter‐cell system has been successfully handled. In this paper, three efficient algorithms have been proposed to reduce the pilot contamination in intra‐cell system. These schemes emphasize on assigning the available pilots to the users based upon their channel qualities by considering the interfering degree of the various pilot sequences. The first proposed algorithm is based on water filling approach, which attains the system throughput very close to that of brute‐force exhaustive search algorithm (ESA) and performs the best among the three proposed algorithms in terms of system throughput. The second algorithm is designed for achieving better user fairness (by assigning less interfering pilots to the poor channel conditioned users). The third algorithm improves the system throughput as compared to the second algorithm without losing much in terms of user fairness. Furthermore, the computational complexity incurred by all the proposed algorithms is much less than that of ESA.

show abstract

“…entries with zero mean and variance σ 2 N . 35 The received user data at the n th BS antenna are given by…”

Section: System Modelmentioning

confidence: 99%

“…Using the matched filter (MF) detector, the received symbol c x up jk can be obtained from the channel estimate d h jkn as given below: 35 c…”

Section: System Modelmentioning

confidence: 99%