The proposed scheme (as illustrated in Fig. 2) was built on the foundation of a standard radio resources management (CRRM) approach, seamlessly integrated into the media independent handover (MIH) architecture, leading to global joint transmission strategies (Irmer et al., 2011; Lee et al., 2012; Wu et al., 2009). In this study, the total bandwidth allocated to the orthogonal frequency division multiplexing (OFDM) scheme was evenly divided into PRB segments. The allocation of PRBs was based on the particular state of each segment. MIMO technology plays a pivotal role in LTE-A, with eNodeBs equipped with multiple antennas and user devices similarly equipped (Chen et al., 2017; Hamrouni et al., 2016; Hamrouni & Hamza, 2017; Hamrouni et al., 2017; Hamrouni et al., 2014). Zero-forcing beamforming ensures efficient data stream transmission for each user (Huh et al., 2011; Somekh et al., 2009).
Mt = ∑ iE = 1 M ei, with E = number of eNodeB.
$${X}_{k}^{p}={Y}_{k}(\sum _{i=1}^{E}{H}_{k,i}^{p}{W}_{k}^{p}{S}_{k, j}^{p}+\sum _{j=1, g \ne k}^{K}\sum _{I=1}^{E}{H}_{k,i}^{p}\sum _{I=1}^{E}{H}_{k,i}^{p} {W}_{j,i}^{p} {S}_{j,i}^{p} +{N}_{k}^{p})$$
1
$${W}_{j,i}^{p}\in {C}^{{M}_{e} X I}$$
\({H}_{j,i}^{p}\in {C}^{{M}_{r} X {M}_{e}}\) : The channel matrix between user k and eNodeB.
\({S}_{j,i}^{p}\in {C}^{{M}_{e} X I}\) : The data vector transmitted by the eNodeB
$$E\left[{S}_{j,i}^{p}{\left({S}_{j,i}^{p}\right)}^{H}\right]=1{N}_{k}^{p}: \text{A}\text{d}\text{d}\text{i}\text{t}\text{i}\text{v}\text{e} \text{w}\text{h}\text{i}\text{t}\text{e} \text{G}\text{a}\text{u}\text{s}\text{s}\text{i}\text{a}\text{n} \text{n}\text{o}\text{i}\text{s}\text{e}.$$
$${SINR}_{k,e}^{p}= \frac{{\left[{Y}_{k}{H}_{k,i}^{p}{W}_{k,e}\right]}^{2}}{\sum _{i=I, i\ne e}^{E}{\left[{Y}_{k}{H}_{k,i}^{p}{W}_{k,e}\right]}^{2}+\sum _{\begin{array}{c}j=1\\ J\ne k\end{array}}^{k}\sum _{i=1}^{E}{\left[{Y}_{k}{H}_{k,i}^{p}{W}_{k,e}\right]}^{2}+{Y}_{k}{\partial }_{k}}$$
2
However, with Joint Transmission, presented in Fig. 3.
In practice, inter-user interference monitoring is an impossible task (Zhang et al., 2023). We operate with a strategy based on the limited threshold value. \({\mu }_{\theta }\)> 0.
$$\begin{array}{cc}\sum _{\begin{array}{c}j=1\\ j\ne k\end{array}}^{E}\sum _{i=I}^{E}{\left|{Y}_{k}{H}_{k,i}^{p}{W}_{j,i}^{p}\right|}^{2}\le & {\mu }_{\theta }\end{array}$$
3
According to the description given above, the SINR represents the Signal to Interference Noise Ratio (Lim et al., 2023), whereas, in information theory and telecommunication engineering, the SINR quantity used to give theoretical upper bounds on channel capacity in wireless communication systems, which at user k formulated as:
$${SNR}_{k,e}^{p}= \frac{{\left|{Y}_{k}{H}_{k,i}^{p}{W}_{j,i}^{p}\right|}^{2}}{\sum _{\begin{array}{c}j=1\\ j\ne k\end{array}}^{k}\sum _{i=I}^{E}{\left|{Y}_{k}{H}_{k,i}^{p}{W}_{j,i}^{p}\right|}^{2}+ {\left]{Y}_{k}{\partial }_{k}\right]}^{2} {⌈{Y}_{k}{\partial }_{k}⌉}^{2}}$$
4
$$\ge \frac{{\left|{Y}_{k}{H}_{k,e}^{p}{W}_{k,e}\right|}^{2}}{{\mu }_{\theta } +{⌈{Y}_{k}{\partial }_{k}⌉}^{2}}$$
$${R}_{k,e}=\frac{I}{{N}_{p}}\sum _{p=1}^{{N}_{p}}{\text{log}}_{2}(1+{SNR}_{k,e}^{p})$$
5
$${C}_{T}=\sum _{e=I}^{E}\sum _{k=I}^{K}{R}_{k,e}$$
6
2.1. Inter-Mobile Users’ Cooperation
In practice, we classify users in edge, and one is selected as a central element. We collect them as groups to constitute an MU team based on three metrics: localization, user’s application, and user’s velocity. Based on an agreement with the operator, we assign a team leader role to a UE. This related role can be delegated to another once the principal is getting down for any operating factor, and thus, to maintain the MU team in Fig. 4.
In the cell, the MU team leader will control the signalization traffic and decrease its value so its radio resource is continuously communicated to the eNodeB.
2.2. PRB Allocation Optimization
The objective of the current study is to ensure an efficient radio resource selection allocated to each MUE team, so all users in the MUE team must be determined and required PRBs estimated. The traffic type influences the required quality of service for each user in the MU team (Phan et al., 2016). Real-time and non-real-time traffic are both controlled. Therefore, for that reason, real-time user applications operate at a fixed rate. Whereas, for those under nonreal-time services, a minimum rate is compulsory. To determine the required PRB number of users, the formula (7) is shown below:
$$⟦{n}_{i}=\frac{{RR}_{i}}{{R}_{PRB}}⟧$$
7
$${R}_{PRB}={M}_{t} X 720 \frac{Kb}{S}$$
8
Then,
$${N}_{k}=\sum _{i}{n}_{i}$$
9
$$\begin{array}{cc}{C}_{1}:\sum _{\begin{array}{c}j=1\\ j\ne k\end{array}}^{E}\sum _{i=I}^{E}{\left|{Y}_{k}{H}_{k,i}^{p}{W}_{j,i}^{p}\right|}^{2}\le & {\mu }_{\theta , }\end{array} \forall i\in \left\{1,.E\right\}, \forall j\in \left\{1,.k\right\}$$
\(\begin{array}{cc}{C}_{2}:{R}_{k,i}^{RT}& =\end{array} {\mu }_{RT} \forall k\in\) {1,.\({K}^{RT}\)}\(\) (10)
\(\begin{array}{cc}{C}_{3}:{R}_{k,i}^{NRT}& \ge \end{array} {\mu }_{NRT} \forall k\in\) {1,.\({K}^{NRT}\)}\(\)
$$\begin{array}{c}{{C}_{4}:C}_{T}\le {C}_{threshold}\end{array}$$
For each MU team, we select a GPRB, and for efficiency maximization throughput, for the GPRB, we downlink the related beamforming vector. The MIIS Media Independent handover Information Server monitors each step to extend the available capacity for heterogeneous cells. All deployed operators in the area, such as macro or small cells, contribute to system stability.