Dynamic priority-based bandwidth allocation scheme for machine type communications

Abstract

The burgeoning realm of Machine Type Communications (MTC) has ushered in a paradigm shift in wireless communication, necessitating innovative strategies to efficiently manage diverse application requirements. This paper presents a pioneering Dynamic Priority-based Bandwidth Allocation (DPBA) scheme tailored specifically for MTC scenarios, encompassing both MTC and Human Type Communications (HTC) coexistence. DPBA employs an adaptive framework addressing dynamic MTC traffic, optimizing resource use while meeting latency constraints and sporadic data patterns. By dynamically prioritizing MTC applications based on requirements, the scheme allocates bandwidth for reliable data delivery. DPBA extends to MTC and HTC coexistence, managing diverse quality demands. This underlines its versatility and capacity to serve distinct needs in a shared spectrum. To substantiate its performance, the DPBA scheme is rigorously evaluated against a spectrum of related allocation methods like Proportional Fairness (PF), Static Priority Scheduling (SPS), and the Machine Learning-based Scheme (MLS), DPBA excels in simulations, minimizing packet loss, latency, enhancing throughput and ensuring fairness, surpassing benchmarks. This research underscores the critical importance of tailored bandwidth allocation strategies in advancing MTC and HTC coexistent environments. With its adaptability, efficiency, and remarkable performance, the DPBA scheme emerges as a cornerstone in the trajectory of wireless communication networks, catering to the distinct requirements of both MTC and HTC applications.

1 Introduction

The proliferation of cellular network technologies has led to the management of an extensive array of interconnected devices becoming increasingly complex [1]. With the exponential growth in online users and devices, driven by the widespread availability of the internet and the emergence of the Internet of Things (IoT), the significance of Machine Type Communications (MTC) has surged across various domains [3] and forecasts predict over 30 billion IoT connections by 2032 [17], as depicted in Fig. 1.

Fig. 1

IoT Connections Forecasts. Bar graph shows the increasing IoT connections (in billions) forecast from 2021 to 2032

1.1 Motivation

The realm of MTC involves the coordination of a multitude of wireless devices coexisting with Human Type Communications (HTC) [18]. MTC’s potential to enable innovative business practices and unlock opportunities is a driving force. These devices find applications in diverse IoT contexts, reshaping communication paradigms. The unique attributes of MTC traffic dynamics, including uplink concentration, sporadic data transmission, diverse Quality of Service (QoS) requirements, mobility, and constrained resources [3], necessitate tailored management to optimize performance across a spectrum of applications.

1.2 Challenges in MTC and HTC coexistence

The coexistence of MTC and HTC in modern wireless communication systems presents several significant challenges and these include:

• 1. Diverse QoS Requirements: MTC and HTC have fundamentally different QoS requirements. MTC often demands real-time data transmission with minimal delay and high reliability; while, HTC focuses on maximizing throughput and maintaining an acceptable level of latency. Balancing these conflicting needs within a single network is a complex task.

• 2. Scalability Issues: The sheer number of MTC devices, which can be orders of magnitude higher than HTC devices, introduces scalability challenges. Networks must efficiently manage a vast number of small data transmissions without compromising the service quality of HTC applications.

• 3. Resource Allocation Efficiency: Traditional resource allocation methods, such as Proportional Fairness (PF) and Maximum- Largest Weighted (MLW), are optimized for HTC and do not account for the sporadic and bursty nature of MTC traffic. This inefficiency often results in suboptimal use of available bandwidth, leading to performance degradation.

• 4. Network Congestion: The simultaneous transmission of MTC and HTC can lead to network congestion, causing increased latency and packet loss. Effective congestion management strategies are essential to maintain high network performance and QoS.

1.3 Potential value of DPBA solutions

The Dynamic Priority-based Bandwidth Allocation (DPBA) scheme proposed in this paper aims to address these challenges by dynamically adjusting the priority of MTC applications based on real-time network conditions. The key innovations and potential benefits of the DPBA scheme include:

• 1. Dynamic Priority Adjustment: By dynamically adjusting the priority of MTC traffic, DPBA ensures that critical MTC data are transmitted with minimal delay, enhancing reliability and responsiveness.

• 2. Improved QoS Balance: DPBA effectively balances the QoS requirements of both MTC and HTC, optimizing bandwidth allocation to meet the diverse needs of both types of traffic. This balance ensures that neither type of communication is unduly prioritized over the other.

• 3. Scalability and Flexibility: The DPBA scheme is designed to scale with the increasing number of MTC devices and adapt to varying traffic patterns. This scalability ensures that the network can efficiently handle large volumes of MTC traffic without compromising HTC performance.

• 4. Enhanced Network Performance: By mitigating congestion and optimizing resource allocation, DPBA improves overall network performance, reducing latency, packet loss, and enhancing throughput. This results in a more reliable and efficient communication system capable of supporting a wide range of applications.

1.4 Organization of the paper

The rest of this paper is organized as follows: Section 2 reviews related work on bandwidth allocation strategies in heterogeneous networks. Section 3 presents the methods including the system model and the DPBA scheme in detail with its algorithmic implementation. Section 4 describes the experimental setup and discusses the results of our experiments, comparing DPBA with existing methods such as PF, SPS, and MLS. Finally, Section 5 concludes the paper with a summary of our findings and potential directions for future research.

2 Literature review

The dynamic landscape of wireless communication networks has prompted research efforts to address the challenges posed by the proliferation of MTC alongside traditional HTC. In this section, we present a comprehensive review of the relevant literature concerning bandwidth allocation schemes in the context of MTC, highlighting their strengths and limitations. The coexistence of MTC and HTC in wireless networks introduces new challenges that demand specialized resource allocation strategies. Machine Learning-based Scheme (MLS) leverage predictive analytics to dynamically allocate resources based on traffic patterns and network conditions [11, 19, 22]. While MLS schemes offer adaptability, they often require substantial training data and may not perform optimally in highly dynamic scenarios. In the realm of HTC/MTC coexistence, radio resources can be categorized into two fundamental approaches: orthogonal and shared allocation. The authors in [16] have introduced these two methods for scheduling MTC uplink transmissions. The sophisticated algorithms make an effort to manage delays and assess the quality of the connection; hence, a need for innovative method for organizing tasks and this was identified in a research investigation covered by the authors in [23]. The researchers in [23] categorized the services into two distinct groups: interpersonal communication (HTC) and automated communication (MTC). They employed diverse strategies to organize their timetable for each group, with a particular focus on ensuring adequate time for individual discussions. The aforementioned studies [16, 23] examined the concept of allocating radio resources uniquely, with a specific emphasis on the algorithms used to determine the scheduling of communication for MTC Devices (MTCDs) and Cellular User Equipment (CUE). A significant issue associated with these algorithms is the potential resource depletion in Multi-Tier Cellular Devices, particularly in scenarios involving high volumes of HTC. In an alternative study, the authors in [14] proposed an intelligent approach to manage the allocation of resources in MTC systems where a technique was devised to categorize comparable entities, thereby streamlining and enhancing effectiveness. Every cluster was assigned a designated time period known as the Access Grant Time Interval (AGTI). The interval was established using a predetermined threshold for timing variation. The primary objective of their research was to examine consistent traffic patterns, as opposed to the unpredictable nature of real MTC. In addition, they failed to recognize the significance of evaluating the channel quality for every device to improve the overall speed of data transmission. The authors in [15], discussed the concept of Access Class Barring (ACB) as a potential solution for addressing the problem of congestion in MTC. It is essential to acknowledge that this concept primarily functions in the realm of access management rather than resource distribution. The primary issue encountered by these works pertains to the challenge of effectively regulating the flow of traffic and efficiently managing the transmission capacity for data. Some resource allocation methodologies focus solely on MTC as with case of Gotsis et al. [9] who performed the analysis of MTC scheduling from a queueing theory standpoint. Their research considered statistical delay requirements while disregarding channel conditions. A study was conducted in [8] by examining group-based MTC resource allocation with objective of classifying MTCDs according to their transmission protocol and QoS requirements. Following this, a joint endeavor was initiated to enhance the distribution of power and resources to maximize the overall output. The researchers in [6] presented an algorithm for resource allocation in event-driven MTC applications within the LTE uplink. This algorithm was designed to provide prognostic capabilities. However, the algorithm lacked resource utilization in multi-channel environments and a diverse range of communication involving MTC and HTC simultaneously. The authors in [2] introduced the TCA algorithm to cater for particular requirements of MTC within the framework of 5G networks whereby in their algorithm, Mobile Traffic Control Devices choose Resource Blocks (RBs) based on their battery level and the power profile of the associated application. On the other hand, Bartoli et al. [5] proposed a two-tiered framework for cluster-based Mobile Traffic Control Devices. In their system, the Mobile Traffic Control Devices send their information to a designated cluster gateway, which then transfers the data to the Base Station (BS). Subsequently, a suggestion was put forth to address a problem in distributing resources to decrease the chances of losing data packets and optimizing the use of available frequencies. The Hybrid Non-Orthogonal Random Access and Data Transmission (NORA-DT) technique was suggested by the authors in [21] as a solution to address the problems associated with signaling overhead and resource allocation in MTC. Also the research conducted by Atchome et al.  [4] examined a hybrid network model that incorporates both MTC and Device-to-Device Communication (D2D). A methodology based on stochastic geometry was used to develop mathematical equations for data transmission rates and assess the overall efficiency of the network in terms of effective throughput.

Based on this scholarly literature reviewed, it is evident that efficient bandwidth allocation has emerged as a critical factor in optimizing the performance of MTC applications. The allocation process must accommodate the distinctive characteristics of MTC traffic dynamics, which often involve sporadic bursts of data transmission and stringent latency constraints. PF schemes have gained attention for their ability to provide equitable resource sharing among diverse MTC applications [13]. PF schemes allocate bandwidth based on a fairness metric that balances the transmission rates of different users or applications. However, PF schemes may not fully capture the varying priorities of MTC applications, leading to suboptimal resource utilization in scenarios with distinct QoS requirements. Another notable approach is the Static Priority Scheduling (SPS) strategy, which assigns static priorities to different MTC applications based on their requirements  [7, 10, 20]. While SPS improves resource allocation efficiency, it lacks adaptability to changes in network conditions and the coexistence of HTC traffic. Recognizing the limitations of static allocation approaches, recent studies have explored dynamic resource management strategies for MTC although with limited exploration of MTC/HTC coexistence. Therefore, the proposed Dynamic Priority-Based Bandwidth Allocation (DPBA) scheme stands as an innovative solution to the challenges posed by the coexistence of MTC and HTC. DPBA dynamically adjusts MTC Real-Time (MTC-RT) arrival rates based on fluctuations in HTC communication load, ensuring efficient bandwidth sharing while maintaining QoS requirements. By adapting to real-time network conditions, DPBA optimizes resource allocation and enhances the overall network performance.

To evaluate the efficacy of DPBA, a comparative analysis is conducted against existing schemes including PF, SPS, and MLS. While PF and SPS address some aspects of resource allocation, they fall short in scenarios with dynamic traffic patterns and coexistence requirements. MLS schemes leverage machine learning, but their performance depends on the quality and quantity of training data.

In summary, the existing literature underscores the significance of tailored resource allocation strategies for optimizing the performance of MTC applications in the presence of HTC traffic. The proposed DPBA scheme introduces a dynamic approach that responds to real-time changes, striking a balance between efficient resource utilization and meeting diverse QoS demands. In the subsequent sections, we present the problem formulation and evaluation results of the DPBA scheme, highlighting its superiority over existing methods.

3 Methods

In this section, we present a comprehensive problem formulation for the proposed DPBA scheme. We detail the mathematical model and derive the dynamic allocation mechanism that addresses the challenges of coexisting MTC and HTC within Long Term Evolution (LTE) networks.

3.1 System model

The system model used to evaluate the performance of the DPBA scheme within the context of LTE networks coexisting with both MTC and HTC considers a cellular network scenario with a single base station (eNodeB) serving a heterogeneous set of users, including both MTCDs and HTC Devices (HTCDs) as shown in Fig.2. The eNodeB operates within a carrier frequency of 2.5 GHz and a bandwidth of 20 MHz.

Fig. 2

Proposed System Model. The model consists of core network and radio access network with the Base Station serving both MTCDs and HTCDs through the radio links

For the simulation, we employ a combination of exponential and Poisson traffic models to mimic the diverse communication behaviors of MTC and HTC applications, respectively. The arrival rate of HTC devices, denoted as $\lambda _{\text{HTC}}$, varies from 15 to 60 with intervals of 5, while the arrival rate of MTC devices, denoted as $\lambda _{\text{MTC}}$, ranges from 25 to 70 with intervals of 5. Each MTCD and HTCD generates data packets according to their respective traffic models. MTCDs transmit real-time (RT) and non-real-time (NRT) traffic, while HTCDs transmit conventional data. The packets are then queued for transmission.

The DPBA scheme dynamically assigns priorities to the incoming traffic based on the system’s objectives and the fluctuating load of HTC traffic. The priority parameters $\alpha$, $\beta _m$, and $\epsilon$ are carefully tuned to achieve the desired trade-off between maximizing throughput, ensuring timely delivery, and maintaining fairness.

Once the priorities are assigned, the DPBA scheme allocates available bandwidth to the queued packets. The bandwidth allocation is governed by the optimization problem formulated to achieve the scheme’s objectives. The Lagrange Multiplier method is employed to find optimal values of parameters that maximize throughput while satisfying latency and fairness constraints.

We evaluate the DPBA scheme’s performance using key metrics including throughput, latency, fairness, and packet loss rate. These metrics provide insights into the efficiency, reliability, and fairness of the DPBA scheme in coexisting MTC and HTC scenarios.

The system model presented herein provides the foundation for evaluating the effectiveness of the DPBA scheme. By incorporating diverse traffic models, dynamic priority assignment, and optimized bandwidth allocation, the DPBA scheme addresses the challenges posed by the coexistence of MTC and HTC in LTE networks. The subsequent sections delve into priority assignment and adjustment, resource allocation, mathematical modeling and algorithmic design, and presentation of the performance metrics.

3.2 Dynamic priority-based bandwidth allocation

3.2.1 MTC priority calculation

The DPBA scheme involves assigning priority levels to MTC applications based on their specific requirements and the current HTC communication load. We represent the set of MTC applications as $\mathcal {M}$, and each application $m \in \mathcal {M}$ is assigned a priority level $p_m$. The priority is calculated as follows:

$$p_m = f({\text{MTC Characteristics}}, {\text{HTC Load}})$$

(1)

where $f(\cdot )$ is a function that takes into account the unique characteristics of each MTC application and the current load on the HTC communication.

3.2.2 Dynamic MTC real-time arrival rate adjustment

The DPBA scheme dynamically adjusts the MTC-RT arrival rate based on the priorities assigned to MTC applications and the current HTC communication load. The objective is to optimally allocate bandwidth while preserving the QoS for both MTC and HTC applications.

The adjusted MTC-RT arrival rate $\lambda _{m,\text {adjusted}}$ for an MTC application m is calculated as:

$$\begin{aligned} \lambda _{m,\text {adjusted}} = \lambda _{m,\text {initial}} \times \left( 1 - \frac{p_m}{\sum _{m' \in \mathcal {M}} p_{m'}}\right) \end{aligned}$$

(2)

where: $\lambda _{m,\text {initial}}$ is the initial arrival rate for application m. m represents an individual MTC application for which the adjusted arrival rate is being calculated. $m'$ represents an individual MTC application within the set $\mathcal {M}$ of all MTC applications. $\mathcal {M}$ denotes the set of all MTC applications. $p_m$ is the priority of MTC application m, determined by the dynamic priority calculation mechanism.

3.2.3 Mathematical model

The primary objective of the DPBA scheme is to maximize the overall throughput while ensuring that both MTC and HTC applications receive timely and reliable data delivery. The problem can be formulated as follows:

$$\begin{aligned} & {\text{Maximize}}\quad \sum\limits_{{m \in {\mathcal{M}}}} {\lambda _{{m,{\text{adjusted}}}} } \times {\text{Throughput}}_{m} \\ & {\text{Subject to:}} \\ & \quad \sum {m \in {\mathcal{M}}} \lambda _{{m,{\text{adjusted}}}} \times {\text{Throughput}}_{m} \le B \\ & \;\quad {\text{Latency}}_{m} \le {\text{MaxLatency}}_{m} \\ & \;\quad {\text{QoS}}_{m} \ge {\text{MinQoS}}_{m} \\ & \;\quad {\text{Fairness}}_{m} \ge {\text{MinFairness}}_{m} \; \\ \end{aligned}$$

(3)

Where the constraints are focusing on bandwidth allocation, latency, QoS and fairness.

Optimization Problem Formulation

Since the primary objective of the DPBA scheme is to maximize the overall throughput while ensuring timely and reliable data delivery for both MTC and HTC applications, while also promoting fairness among different applications, the optimization problem can be reformulated as follows:

$$\begin{aligned} \text {Maximize} \quad \sum _{m \in \mathcal {M}_{\text {RT}}} R_m \end{aligned}$$

(4)

Subject to:

• 1.
  $$\sum _{m \in \mathcal {M}_{\text {RT}}} R_m \le B_{\text {total}}$$
• 2.
  $$R_m \le \lambda _{m,\text {adjusted}} \times p_m$$
• 3.
  $$\min _{m_1, m_2 \in \mathcal {M}_{\text {RT}}, m_1 \ne m_2} |R_{m_1} - R_{m_2}| \le \epsilon$$

where $R_m$ represents the data rate of MTC-RT application $m$, and $\mathcal {M}_{\text {RT}}$ is the set of all MTC-RT applications, $\lambda _{m,\text {adjusted}}$ is the adjusted arrival rate for application $m$ and $p_m$ is its priority, and $\epsilon$ is a small positive constant that controls the degree of fairness. The first constraints indicates that the total allocated bandwidth should not exceed the available total bandwidth $B_{\text {total}}$. The second constraint says that the allocated bandwidth for each MTC-RT application $m$ should not exceed its adjusted arrival rate times its priority. The third constraint promotes fairness among different applications by ensuring that the allocated data rates are distributed as evenly as possible while taking into account their priorities and adjusted arrival rates; hence, this constraint aims to minimize the variation in the allocated data rates.

Solution

The solution to this optimization problem involves determining the data rates $R_m$ for all MTC-RT applications $m$ that satisfy the given constraints while maximizing the overall system throughput and ensuring fairness. The Lagrange Multiplier method is employed to find the optimal values of $R_m$, $\alpha$, $\beta _m$, and $\epsilon$ that achieve the objectives and constraints of the DPBA scheme. We start with the Lagrangian function for the optimization problem:

$$\begin{aligned} \begin{aligned} L&= \sum _{m \in \mathcal {M}{\text {RT}}} R_m - \alpha \left( \sum {m \in \mathcal {M}{\text {RT}}} R_m - B{\text {total}} \right) \ \\&\quad - \sum _{m \in \mathcal {M}{\text {RT}}} \beta _m \left( R_m - \lambda {m,\text {adjusted}} \times p_m \right) \ \\&\quad + \epsilon \sum _{m_1, m_2 \in \mathcal {M}{\text {RT}}, m_1 \ne m_2} \left| R{m_1} - R_{m_2} \right| \end{aligned} \end{aligned}$$

(5)

where $\alpha$ is the Lagrange multiplier for the total bandwidth constraint. $\beta _m$ is the Lagrange multiplier for the bandwidth allocation constraint for each MTC-RT application m. $\epsilon$ is the Lagrange multiplier for the fairness constraint.

We differentiate L with respect to each variable and set the derivatives equal to zero.

Differentiate L with respect to $R_m$ :

$$\begin{aligned} \frac{\partial L}{\partial R_m} = 1 - \alpha - \beta _m + \epsilon \sum _{m_2 \ne m} \text {sign}(R_m - R_{m_2}) \end{aligned}$$

(6)

Setting the derivative equal to zero:

$$\begin{aligned} 1 - \alpha - \beta _m + \epsilon \sum _{m_2 \ne m} \text {sign}(R_m - R_{m_2}) = 0 \end{aligned}$$

(7)

Solving for $R_m$:

$$\begin{aligned} R_m = \frac{\alpha + \beta _m}{1 + \epsilon \cdot \text {count}(m)} \end{aligned}$$

(8)

Where $\text {count}(m)$ represents the number of MTC-RT applications excluding application m.

Differentiate L with respect to $\alpha$:

$$\begin{aligned} \frac{\partial L}{\partial \alpha } = \sum _{m \in \mathcal {M}{\text {RT}}} R_m - B{\text {total}} \end{aligned}$$

(9)

Setting the derivative equal to zero:

$$\begin{aligned} \sum _{m \in \mathcal {M}{\text {RT}}} R_m - B{\text {total}} = 0 \end{aligned}$$

(10)

Differentiate L with respect to $\beta _m$:

$$\begin{aligned} \frac{\partial L}{\partial \beta _m} = -\left( R_m - \lambda _{m,\text {adjusted}} \times p_m \right) \end{aligned}$$

(11)

Setting the derivative equal to zero:

$$\begin{aligned} -\left( R_m - \lambda _{m,\text {adjusted}} \times p_m \right) = 0 \end{aligned}$$

(12)

Solving for $\beta _m$:

$$\begin{aligned} \beta _m = \lambda _{m,\text {adjusted}} \times p_m \end{aligned}$$

(13)

Differentiate L with respect to $\epsilon$:

$$\begin{aligned} \frac{\partial L}{\partial \epsilon } = \sum _{m_1, m_2 \in \mathcal {M}{\text {RT}}, m_1 \ne m_2} \left| R{m_1} - R_{m_2} \right| \end{aligned}$$

(14)

Solving for the optimal value of $\epsilon$ involves analyzing the cases where $R{m_1} > R_{m_2}$ and $R{m_1} < R_{m_2}$ separately to derive conditions for $\epsilon$ based on those cases.

By solving these equations simultaneously, we obtain the optimal values of $R_m$, $\alpha$, $\beta _m$, and $\epsilon$ that satisfy the objectives and constraints of the DPBA scheme.

3.3 Proposed dynamic priority-based bandwidth allocation (DPBA) algorithm

The detailed steps for the proposed DPBA algorithm in algorithm 1 including initialization, iteration process, and termination conditions are presented below.

The DPBA algorithm aims to dynamically allocate bandwidth to MTC-RT applications in LTE networks, ensuring efficient resource utilization and meeting diverse QoS requirements.

Initialization

• 1. Input Parameters:

  • Total available bandwidth: $B_{\text {total}}$

  • Adjusted MTC-RT arrival rates: $\lambda _{m,\text {adjusted}}$

  • Priorities of MTC-RT applications: $p_m$

  • Minimum data rate: $T_{\text {min}}$

  • Maximum fairness parameter: $\epsilon _{\text {max}}$

  • Maximum bandwidth multiplier: $\alpha _{\text {max}}$

  • Minimum threshold for MTC-RT arrival rate: $\lambda _{m,\text {min}}$

• 2. Initialization of Bandwidth Allocation:

  • Call ‘InitializeBandwidth’ function with $\lambda _{m,\text {adjusted}}, p_m, T_{\text {min}}, \lambda _{m,\text {min}}$ to set initial bandwidth allocations $R_m$ for each MTC-RT application.

  • Example: $R_m \leftarrow \text {InitializeBandwidth}(\lambda _{m,\text {adjusted}}, p_m, T_{\text {min}}, \lambda _{m,\text {min}})$

• 3. Set Convergence Flag:

  • Initialize the convergence flag as False: $\text {converged} \leftarrow \text {False}$

Iteration process

• 1. While Loop (Until Convergence): The algorithm iterates until the convergence criteria are met.

• 2. Calculate Total Allocated Bandwidth: Sum the allocated bandwidths for all MTC-RT applications: $R_{\text {total}} \leftarrow \sum _{m} R_m$

• 3. Calculate Lagrange Multiplier $\alpha$: Calculate $\alpha$ using ‘CalculateAlpha’ to ensure the total allocated bandwidth does not exceed $B_{\text {total}}$: $\alpha \leftarrow \text {CalculateAlpha}(R_{\text {total}}, B_{\text {total}})$

• 4. Calculate Lagrange Multiplier $\beta _m$ for Each MTC-RT Application: For each MTC-RT application $m$, calculate $\beta _m$ using ‘CalculateBeta’: $\beta _m \leftarrow \text {CalculateBeta}(R_m, \lambda _{m,\text {adjusted}}, p_m)$

• 5. Calculate Fairness Parameter $\epsilon$: Calculate the fairness parameter $\epsilon$ using ‘CalculateFairness’: $\epsilon \leftarrow \text {CalculateFairness}(R_m)$

• 6. Update Bandwidth Allocation for Each MTC-RT Application: For each MTC-RT application $m$, update the bandwidth allocation based on the calculated $\alpha , \beta _m,$ and $\epsilon$ using ‘UpdateBandwidth’: $R_m \leftarrow \text {UpdateBandwidth}(R_m, \alpha , \beta _m, \epsilon )$

• 7. Check Convergence: Check if the algorithm has converged using ‘CheckConvergence’: $\text {converged} \leftarrow \text {CheckConvergence}(R_m, \alpha , \beta _m, \epsilon )$

Termination conditions

• The algorithm terminates when the convergence flag is set to True, indicating that the bandwidth allocations $R_m$, the Lagrange multipliers $\alpha$ and $\beta _m$, and the fairness parameter $\epsilon$ have stabilized and no further significant changes occur.

Function definitions

• 1. InitializeBandwidth: Initialize bandwidth allocations based on arrival rates, priorities, minimum data rate, and arrival rate threshold.

• 2. CalculateAlpha: Calculate the Lagrange multiplier $\alpha$ to ensure the total allocated bandwidth meets the constraint.

• 3. CalculateBeta: Calculate the Lagrange multiplier $\beta _m$ to enforce bandwidth allocation constraints for each MTC-RT application.

• 4. CalculateFairness: Calculate the fairness parameter $\epsilon$ based on the current bandwidth allocations.

• 5. UpdateBandwidth: Update the bandwidth allocation for each MTC-RT application using the calculated multipliers and fairness parameter.

• 6. CheckConvergence: Determine if the algorithm has converged by checking the stability of $R_m, \alpha , \beta _m,$ and $\epsilon$.

Algorithm 1

Dynamic Priority-Based Bandwidth Allocation (DPBA)

3.4 Performance metrics

The following metrics are used to evaluate the performance of the proposed DPBA scheme;

3.4.1 Throughput

Throughput measures the total data rate achieved by all MTC and HTC applications in the network. In the DPBA scheme, the throughput reflects the amount of data successfully transmitted within a given time frame. It indicates how effectively the available bandwidth is being utilized to carry data. Total Throughput ($T_{\text {total}}$) is calculated as the sum of the throughputs of MTC and HTC applications:

$$T_{{{\mathrm{total}}}} = T_{{{\mathrm{MTC}}}} + T_{{{\mathrm{HTC}}}}$$

(15)

Where:

• $T_{\text{MTC}}$ is the throughput of MTC applications.

• $T_{\text{HTC}}$ is the throughput of HTC applications.

3.4.2 Latency

Latency measures the average delay experienced by data packets from the source to the destination. In the DPBA scheme, minimizing latency is crucial, especially for real-time MTC applications that require timely data delivery.

Average Latency ($D_{\text {avg}}$) is calculated as the sum of latencies of all transmitted packets divided by the total number of packets:

$$\begin{aligned} D_{\text {avg}} = \frac{\sum _{i=1}^{N} D_i}{N} \end{aligned}$$

(16)

Where:

• N is the total number of transmitted packets.

• $D_i$ is the latency of the ith packet.

3.4.3 Packet loss rate

Packet loss rate quantifies the proportion of data packets that do not reach their destination. In the DPBA scheme, minimizing packet loss is essential for reliable data delivery. Packet Loss Rate (PLR) is calculated as the ratio of the number of lost packets to the total number of transmitted packets:

$${\text{PLR}} = {\text{ }}\frac{{{\text{Number}}\;{\text{of}}\;{\text{Lost}}\;{\text{Packets}}}}{{{\text{Total}}\;{\text{Number}}\;{\text{of}}\;{\text{Transmitted}}\;{\text{Packets}}}} \times 100$$

(17)

3.4.4 Fairness index

Fairness index measures the fairness of bandwidth allocation among different applications. DPBA aims to allocate bandwidth fairly between MTC and HTC applications based on their priorities and requirements. Jain’s Fairness Index (J) is a common fairness metric [12] that is calculated as the ratio of the squared sum of throughputs to the sum of the squared throughputs:

$$J = \frac{{\left( {\sum\nolimits_{{i = 1}}^{N} {T_{i} } } \right)^{2} }}{{N \times \sum\nolimits_{{i = 1}}^{N} {T_{i}^{2} } }}$$

(18)

Where:

• N is the number of applications.

• $T_i$ is the throughput of the ith application.

4 Results and discussion

In this section, we present a comprehensive performance analysis of the DPBA scheme. The DPBA scheme was evaluated through extensive simulations and compared against three existing allocation methods, including Proportional Fairness (PF), Static Priority Scheduling (SPS), and Machine Learning-based Scheme (MLS). The simulation parameters used are shown in Table 1.

Table 1 Simulation parameters

4.1 Simulation setup

In order to comprehensively evaluate the performance of the Dynamic Priority-Based Bandwidth Allocation (DPBA) scheme, we conducted simulations using MATLAB. The simulation environment was designed to mimic real-world LTE network conditions and interactions between MTC and HTC applications.

The simulated network topology consisted of a heterogeneous mix of MTC and HTC devices communicating with an LTE base station (eNodeB). The MTC devices represented various IoT applications with diverse data requirements, while the HTC devices represented traditional human communication traffic.

To emulate realistic traffic scenarios, we employed a combination of traffic models for both MTC and HTC applications. MTC applications, which often involve sporadic and bursty data transmissions triggered by specific events or sensor readings, are well-suited to be represented by an exponential traffic model. In this model, data packets are generated with a distribution that reflects the varying time intervals between successive packet arrivals. The exponential distribution captures the unpredictable and intermittent nature of MTC traffic, allowing the DPBA scheme to simulate the abrupt surges in data transmission that characterize many IoT applications. In contrast, HTC applications typically exhibit more regular and continuous data transmission patterns. For this reason, a Poisson traffic model is employed to represent the arrival of data packets from HTC devices. The Poisson distribution characterizes events that occur at a constant average rate but in a seemingly random manner. This model aligns well with human communication behaviors, such as voice calls or messaging, where data packets are generated independently at irregular intervals. By incorporating the exponential traffic model for MTC and the Poisson traffic model for HTC, the DPBA scheme gains the ability to realistically simulate the diverse traffic patterns of both application types within the LTE network. This nuanced approach enables the scheme to dynamically allocate bandwidth based on the unique characteristics of MTC and HTC traffic, optimizing resource utilization, minimizing latency, and ensuring fairness.

To assess the DPBA scheme’s performance across different scenarios, we varied parameters such as the number of MTC and HTC devices, their data rates, and priority levels. Additionally, we introduced dynamic changes in traffic load and application priorities to observe the scheme’s adaptability. Statistical methods were applied to analyze simulation results, including mean values, standard deviations, and confidence intervals, to ensure the validity and reliability of the findings.

The simulation setup aimed to replicate real-world LTE network conditions and interactions between MTC and HTC applications. By varying parameters, introducing dynamic scenarios, and employing statistical analyses, we obtained reliable insights into the performance of the DPBA scheme under diverse conditions.

4.2 Results and discussion

We present the results of our performance evaluation in comparison with PF, SPS and MLS. The following are the key findings:

4.2.1 Throughput

The results of the performance evaluation of the Dynamic Priority-Based Bandwidth Allocation (DPBA) scheme, along with comparison to PF, SPS and MLS shown in Figs. 3 and 4 , reveal valuable insights into the scheme’s effectiveness.

Fig. 3

Throughput Comparison with $\lambda _{\text{HTC}} = 15$ and $\lambda _{\text{MTC}} = 25.$ Blue square shows DPBA, Red circle for PF, Green diamond for SPS, Purple cross for MLS performance

In 1000 iterations, each characterized by HTC traffic load ($\lambda _{\text{HTC}} = 15$) and MTC traffic load ($\lambda _{\text{MTC}} = 25$), the DPBA scheme demonstrates consistent throughput performance. The observed throughput values exhibit fluctuations ranging from 14 Mbps to 20 Mbps. The average throughput achieved by the DPBA scheme across all iterations is 15.71 Mbps. This robust performance is indicative of the DPBA’s dynamic allocation mechanism’s ability to effectively distribute available bandwidth resources while accommodating both MTC and HTC traffic requirements. Contrastingly, the MLS exhibits a different throughput profile. Over the same 1000 iterations, the MLS scheme’s throughput experiences fluctuations between 11 Mbps and 14 Mbps. The average throughput achieved by MLS stands at 12.08 Mbps. This variance in throughput suggests that while MLS may offer some adaptability, it struggles to consistently match the performance of DPBA in scenarios characterized by mixed MTC and HTC traffic. The SPS strategy, which allocates bandwidth based on predetermined priority levels, showcases a consistent throughput value of 10 Mbps across all 1000 iterations. This static approach to bandwidth allocation fails to capitalize on the potential for adaptability and optimization offered by dynamic schemes like DPBA. Consequently, SPS lags behind DPBA in terms of throughput performance. The PF approach, designed to balance fairness and throughput, exhibits the lowest throughput performance among the compared schemes. With throughput fluctuations ranging between 0 and 1 Mbps, the PF scheme struggles to provide satisfactory data delivery rates. The average throughput achieved by PF is merely 0.39 Mbps, highlighting its limitations in scenarios where dynamic adaptation to changing traffic conditions is crucial.

The analysis of throughput in the context of increasing HTC and MTC traffic loads reveals noteworthy trends and differences in the performance of the DPBA scheme as well as the comparison schemes (PF, SPS, MLS) as shown in Fig. 4.

Fig. 4

Throughput Comparison with increasing arrival rates for HTC ($\lambda _{\text{HTC}} =15:5:60$) and MTC ($\lambda _{\text{MTC}} = 25:5:70$). Blue square shows DPBA, Red circle for PF, Green diamond for SPS, Purple cross for MLS performance

As both HTC ($\lambda _{\text{HTC}}$) and MTC ($\lambda _{\text{MTC}}$) traffic loads increase, the impact on throughput becomes evident. In a series of 1000 iterations, where $\lambda _{\text{HTC}}$ ranges from 15 to 60 in intervals of 5, and $\lambda _{\text{MTC}}$ ranges from 25 to 70 in intervals of 5, the throughput values of DPBA, PF, SPS, and MLS schemes exhibit varying trends. The DPBA scheme’s throughput experiences a gradual decline as both HTC and MTC traffic loads increase. The throughput values for DPBA decrease from 15.71 to 12.30 Mbps, reflecting the scheme’s adaptability to dynamically allocate bandwidth while ensuring efficient coexistence of MTC and HTC traffic. Despite the decrease, DPBA maintains competitive throughput values compared to the other schemes. The PF approach showcases a more pronounced decrease in throughput as traffic loads increase. PF’s throughput values decline from 0.39 to 0.01 Mbps, indicating its limited ability to balance fairness and throughput under changing traffic conditions. The decreasing trend highlights PF’s challenges in handling the dynamics of both MTC and HTC traffic. SPS exhibits a consistent and gradual decrease in throughput as traffic loads increase. The throughput values for SPS decrease from 10.00 to 5.00 Mbps. The static nature of SPS hinders its ability to effectively adapt to varying traffic scenarios, resulting in a linear decline in throughput. The MLS experiences a relatively stable decrease in throughput as traffic loads increase. MLS’s throughput values decline from 12.08 to 6.80 Mbps. While MLS demonstrates some adaptability, its performance deteriorates more than the DPBA by at least 40% as the complexity of traffic conditions intensifies. In summary, the analysis of throughput under stable and increasing HTC and MTC traffic loads reveals that the DPBA scheme maintains competitive throughput values, showcasing its adaptability and optimization capabilities. PF and SPS suffer from decreased throughput due to their static nature, while MLS experiences a more gradual decline. These findings highlight DPBA’s ability to dynamically allocate bandwidth resources and optimize performance under varying traffic conditions, making it a promising choice for managing the coexistence of MTC and HTC applications within LTE networks. The comparison to other schemes underscores the advantages of the DPBA approach in optimizing bandwidth allocation, resulting in improved data delivery rates and overall network efficiency.

4.2.2 Latency

The examination of latency within different iterations and traffic conditions provides insights into the performance of the DPBA scheme as compared to the related schemes PF, SPS, and MLS. The simulation results are shown in Fig. 5.

Fig. 5

Latency Comparison with $\lambda _{\text{HTC}} = 15$ and $\lambda _{\text{MTC}} = 25.$ Blue square shows DPBA, Red circle for PF, Green diamond for SPS, Purple cross for MLS performance

In the 1000 iterations, where both HTC ($\lambda _{\text{HTC}}$) and MTC ($\lambda _{\text{MTC}}$) traffic loads remain constant at 15 and 25; respectively, the DPBA scheme consistently maintains a stable latency of 0.1 ms. This indicates DPBA’s ability to effectively manage latency even under different iteration counts, reflecting its capability to allocate resources dynamically and ensure timely data delivery. The average latency of 0.1 ms across all iterations reinforces DPBA’s capacity to provide low and consistent latency for both MTC and HTC applications. Similar to DPBA, the MLS also exhibits a stable latency of 0.14 ms across all iterations. This stability suggests that MLS is effective in maintaining a relatively consistent latency level, albeit slightly higher than DPBA. SPS consistently maintains a latency of 0.12 ms in all iterations. This too suggests that SPS is effective in maintaining a relatively consistent latency level, albeit slightly higher than DPBA just like MLS. On the other hand, the PF scheme’s latency varies significantly across different iterations. It fluctuates between 0 and 230 ms, with an average latency of 46.79 ms. The wide fluctuation highlights PF’s challenge in providing consistent and low latency, particularly in scenarios with varying traffic conditions. The higher average latency suggests that PF may prioritize fairness over low latency, leading to higher latencies in certain situations.

The investigation of latency across a range of iterations and varying traffic conditions provides significant insights into the performance of the DPBA scheme in comparison with the PF, SPS, and MLS schemes. The results plot is shown in Fig. 6.

Fig. 6

Latency Comparison with increasing arrival rates for HTC ($\lambda _{\text{HTC}} =15:5:60$) and MTC ($\lambda _{\text{MTC}} = 25:5:70$). Blue square shows DPBA, Red circle for PF, Green diamond for SPS, Purple cross for MLS performance

In Fig. 6, it is evident that in a series of 1000 iterations, where both HTC ($\lambda _{\text{HTC}}$) and MTC ($\lambda _{\text{MTC}}$) traffic loads progressively increase in intervals of 5, DPBA’s latency also experiences an upward trend. The DPBA latency values rise as follows: 0.10 ms, 0.26 ms, 0.38 ms,  0.51 ms, 1.30 ms, 2.35 ms,  3.5 ms, 4.56 ms,  5.60 ms, and 6.80 ms. This increase in latency is expected as the network experiences higher traffic loads, and DPBA strives to allocate resources efficiently to accommodate both MTC and HTC applications. The gradual increase in latency is indicative of DPBA’s ability to adapt to changing traffic conditions while maintaining relatively low and acceptable latency levels. In the same set of 1000 iterations, the PF scheme exhibits a substantial increase in latency. The PF latency values increase as follows: 46.79 ms, 49.56 ms,  55.30 ms, 65.25 ms, 70.21 ms,  75.15 ms, 79.61 ms,  83.68 ms, 88.05 ms,  and 96.41 ms. The significant increase in PF’s latency values highlights its challenge in maintaining low and consistent latency as traffic loads increase. PF’s focus on fairness can lead to increased latency, particularly when the network becomes congested. The SPS also experiences a notable increase in latency across the 1000 iterations. The SPS latency values increase as follows: 0.12 ms, 0.70 ms,  1.7 ms, 3.90 ms,  5.80 ms, 6.50 ms, 9.21 ms,  11.92 ms, 14.89 ms,  and 19.90 ms. Similar to PF, SPS’s latency values rise as the network becomes more congested, highlighting its limitations in adapting to changing traffic conditions and maintaining low latency. The MLS also demonstrates an increase in latency over the 1000 iterations. The MLS latency values increase as follows: 0.14 ms, 0.6 ms,  0.9 ms, 2.90 ms,  3.11 ms, 4.50 ms, 6.88 ms,  8.60 ms, 9.67 ms,  and 11.23 ms. The rise in MLS latency values indicates its response to growing traffic loads. While MLS maintains relatively lower latency compared to PF and SPS, it still experiences an increase in latency with congestion more than DPBA.

In summary, the analysis of latency across constant and increasing traffic loads and iterations indicates that the DPBA scheme effectively manages latency by adapting to changing network conditions. DPBA’s latency increases gradually, maintaining relatively low and acceptable levels even under congested scenarios. PF, SPS, and MLS experience more significant increases in latency, highlighting their limitations in managing latency under varying traffic conditions. DPBA’s ability to provide controlled latency while efficiently allocating resources makes it a promising solution for ensuring timely and reliable data delivery for both MTC and HTC applications.

4.2.3 Packet loss rate

The investigation into packet loss rates across varying iterations and traffic loads provides significant insights into the performance of the DPBA scheme in relation to MLS, SPS, and PF schemes.

Fig. 7

Packet Loss Rate Comparison with $\lambda _{\text{HTC}} = 15$ and $\lambda _{\text{MTC}} = 25.$ Blue square shows DPBA, Red circle for PF, Green diamond for SPS, Purple cross for MLS performance

From Fig. 7, it is evident that in all the 1000 iterations with consistent HTC and MTC traffic loads, the DPBA scheme demonstrates a relatively stable packet loss rate ranging between 0.05 and 0.1, with an average packet loss rate of 0.06. This stable and low packet loss rate indicates DPBA’s ability to effectively manage both MTC and HTC traffic while minimizing data loss. The adaptive nature of DPBA allows it to allocate resources optimally, thereby reducing the likelihood of packet loss. Across the same iterations, the MLS exhibits a packet loss rate that fluctuates between 0 and 0.7. The average packet loss rate of 0.12 suggests that MLS may encounter challenges in adapting to varying traffic loads and efficiently allocating resources to prevent packet loss. This inconsistency in packet loss rates could stem from the limitations of MLS in handling dynamic traffic conditions. The SPS scheme showcases packet loss rates that fluctuate between 0.1 and 1 across the three sets of iterations. The high average packet loss rate of 0.81 indicates that SPS’s static priority allocation mechanism may not effectively manage varying traffic demands and coexisting MTC and HTC applications. The consistently elevated packet loss rates highlight SPS’s limitations in adapting to dynamic scenarios. The PF scheme exhibits a fluctuating packet loss rate between 0 and 1 across all iterations. The high average packet loss rate of 0.88 underscores that PF’s approach to resource allocation based on priority and channel quality might lead to higher packet loss rates, particularly in scenarios with varying traffic loads and resource demands.

Using results shown in Fig. 8 we analyze the packet loss rates across a wide range of iterations and varying traffic loads in an effort to dig some valuable insights into the performance of the DPBA scheme in comparison with MLS, SPS, and PF schemes.

Fig. 8

Packet Loss Rate Comparison with increasing arrival rates for HTC ($\lambda _{\text{HTC}} =15:5:60$) and MTC ($\lambda _{\text{MTC}} = 25:5:70$). Blue square shows DPBA, Red circle for PF, Green diamond for SPS, Purple cross for MLS performance

In 1000 iterations with increasing HTC and MTC traffic loads, the DPBA scheme exhibits an escalating packet loss rate from 0.06 to 0.21. This upward trend suggests that as traffic loads increase, DPBA faces challenges in maintaining a low packet loss rate. However, the gradual increase in packet loss rate indicates DPBA’s ability to adapt to varying traffic conditions and allocate resources based on priorities, even though some packets may experience loss under heavier loads. Across the same iterations, the PF scheme demonstrates a consistently high packet loss rate ranging from 0.88 to 0.94. This trend highlights that PF’s resource allocation based on channel quality and priority might lead to substantial packet losses as traffic loads increase. The elevated and consistently high packet loss rates underscore the limitations of PF in ensuring reliable data delivery under dynamic traffic conditions. The SPS scheme showcases packet loss rates that range from 0.81 to 0.89 in the investigated iterations. This trend suggests that SPS’s static priority assignment struggles to manage packet losses effectively as traffic loads become more substantial. The relatively high and increasing packet loss rates reveal SPS’s limitations in adapting to varying traffic demands. In the same set of iterations, the MLS experiences a packet loss rate increase from 0.12 to 0.38. This trend implies that MLS may face challenges in adapting to changing traffic loads and efficiently allocating resources to prevent packet loss. The increasing packet loss rates highlight potential shortcomings in MLS’s adaptability to dynamic scenarios.

In summary, the analysis of packet loss rates over a range of traffic loads emphasizes that the DPBA scheme showcases a relatively controlled increase in packet loss rates as traffic loads escalate. This indicates DPBA’s capability to dynamically allocate resources and mitigate packet losses to some extent even under heavier loads. In contrast, PF, SPS, and MLS exhibit consistently higher packet loss rates, especially as traffic loads increase, revealing their limitations in handling dynamic traffic conditions.

4.2.4 Fairness index

Based on the results plot shown in Fig. 9, we delve into the findings in relation to the schemes simulated.

Fig. 9

Fairness Comparison with $\lambda _{\text{HTC}} = 15$ and $\lambda _{\text{MTC}} = 25.$ Legend for the ninth figure. Blue square shows DPBA, Red circle for PF, Green diamond for SPS, Purple cross for MLS performance

Across the 1000 iterations, where both HTC ($\lambda _{\text{HTC}}$) and MTC ($\lambda _{\text{MTC}}$) traffic loads remain constant at 15 and 25; respectively, DPBA’s fairness exhibits fluctuations within the range of 0.5 to 0.9. The average fairness achieved by DPBA across these iterations is 0.68. The observed fluctuations in fairness indicate that DPBA dynamically adjusts resource allocation based on changing traffic conditions, aiming to achieve a balanced allocation for both MTC and HTC applications. DPBA’s ability to maintain fairness within this range signifies its capacity to respond to variations in traffic load while still providing a relatively equitable allocation of resources. In the same 1000 iterations, the MLS displays fairness values that fluctuate between 0.3 and 0.7, with an average fairness of 0.56. The observed fairness fluctuations indicate MLS’s effort to achieve fairness, albeit with some variability. However, the fairness values achieved by MLS are lower on average compared to DPBA, suggesting that MLS might struggle to balance resource allocation optimally under dynamic traffic conditions. SPS maintains a consistent fairness value of 0.41 across 1000 iterations. The stability in SPS’s fairness values points out that SPS prioritizes certain types of traffic consistently, potentially leading to underutilization of resources for other traffic types. While SPS may ensure some level of fairness, its lack of adaptability and allocation diversity could hinder its overall performance in dynamic scenarios. The PF scheme maintains a stable fairness value of 0.61 across all iterations. While PF aims to achieve fairness by considering both priority and channel quality, its relatively consistent fairness value might not account for changing traffic dynamics. The consistent fairness value could lead to suboptimal resource allocation, especially when traffic conditions shift rapidly.

Based on Fig. 10 we now look through the analysis of fairness indices across a range of iterations with varying HTC ($\lambda _{\text{HTC}}$) and MTC ($\lambda _{\text{MTC}}$) traffic loads in order to provide valuable insights into the equitable resource allocation performance of the DPBA scheme in comparison with the MLS, SPS, and PF schemes.

Fig. 10

Fairness Comparison with increasing arrival rates for HTC ($\lambda _{\text{HTC}} =15:5:60$) and MTC ($\lambda _{{\text{MTC}}} = 25:5:70$). Blue square shows DPBA, Red circle for PF, Green diamond for SPS, Purple cross for MLS performance

In 1000 iterations, with increasing HTC and MTC traffic loads, the DPBA fairness index experiences a gradual decrease from 0.68 to 0.49. This decreasing trend indicates that DPBA is responsive to changing traffic conditions, as it attempts to maintain fairness while adapting to fluctuations in the traffic load. DPBA’s ability to adjust its resource allocation dynamically showcases its effectiveness in achieving fairness in varying scenarios. Across the same set of iterations, the PF scheme’s fairness index decreases from 0.61 to 0.37. The decreasing trend in fairness indicates that PF’s approach to allocating resources based on priority and channel quality might struggle to maintain equitable allocation as traffic loads increase. While PF aims for proportional fairness, it is not as adaptive as DPBA in handling changing traffic dynamics. The SPS displays a consistent decrease in fairness index from 0.41 to 0.08 as traffic loads increase. The decreasing fairness trend highlights that SPS’s static priority allocation leads to decreasing fairness as traffic loads become more dynamic and diverse. SPS’s lack of adaptability results in reduced fairness when faced with varying traffic conditions. The MLS exhibits a gradual decline in fairness index from 0.56 to 0.23 across the same iterations. The decreasing trend in fairness suggests that MLS’s approach to resource allocation based on machine learning might not adequately address the fairness requirements as traffic loads increase. The decline in fairness values indicates MLS’s limitations in handling the dynamic allocation needed for coexisting MTC and HTC applications.

In summary, the analysis of fairness indices across different traffic load scenarios demonstrates that DPBA maintains relatively high fairness values while adapting to dynamic traffic conditions. As HTC and MTC traffic loads increase, DPBA exhibits a more gradual decline in fairness compared to other schemes. This showcases DPBA’s robustness in maintaining equitable resource allocation, making it a promising solution for addressing fairness challenges in LTE networks with coexisting MTC and HTC applications.

4.3 Significance of experimental results

The experimental results of the DPBA scheme demonstrate its efficacy in managing the coexistence of MTC and HTC within LTE networks. By dynamically adjusting the priorities of MTC applications, the DPBA scheme successfully ensures equitable resource allocation, maintaining the desired service levels for both MTC and HTC traffic. This subsection explores the broader implications of these results and their potential impact on various application scenarios.

4.3.1 Potential impact in different application scenarios

• 1. Smart Cities:

  • Traffic Management: Smart traffic lights, connected vehicles, and public transportation systems rely heavily on real-time data exchange. The DPBA scheme can prioritize critical MTC applications, such as emergency vehicle notifications, while ensuring smooth data flow for HTC applications used by commuters.

  • Public Safety: Surveillance cameras and environmental sensors generate large volumes of data that need to be transmitted reliably and in real-time. DPBA ensures these high-priority MTC applications receive adequate bandwidth, improving response times and situational awareness for emergency services.

• 2. Industrial IoT:

  • Manufacturing: In smart factories, numerous sensors and machines communicate to optimize production processes. The DPBA scheme can prioritize time-sensitive MTC applications (e.g., machine status updates) over less critical HTC applications (e.g., employee communications), ensuring minimal latency and improved operational efficiency.

  • Predictive Maintenance: For predictive maintenance applications, sensors continuously monitor equipment health. DPBA can allocate bandwidth dynamically to these critical MTC applications, enabling timely data transmission and early fault detection, thereby reducing downtime and maintenance costs.

• 3. Healthcare:

  • Remote Patient Monitoring: Wearable devices and sensors continuously collect patient health data. DPBA can prioritize these MTC applications, ensuring reliable and timely data transmission to healthcare providers, thus enhancing patient care and enabling quicker medical responses.

  • Telemedicine: In telemedicine scenarios, video consultations (HTC) and real-time health monitoring (MTC) coexist. DPBA ensures that both types of applications receive appropriate bandwidth, improving the quality of care and patient outcomes.

• 4. Smart Homes: Smart home devices, such as security systems, climate control, and energy management, generate various MTC data streams. DPBA can prioritize critical MTC applications (e.g., security alerts) while maintaining a balanced resource allocation for HTC applications (e.g., streaming services), enhancing the overall user experience.

4.3.2 Challenges in practical deployment

Despite its advantages, the practical deployment of the DPBA scheme may encounter a number of challenges and these include:

• 1. Scalability: As the number of connected devices in a network increases, the computational complexity of dynamically adjusting priorities and bandwidth allocations may become a bottleneck. Efficient algorithms and optimization techniques will be required to ensure scalability without compromising performance.

• 2. Dynamic Network Conditions: Real-world networks experience varying traffic loads and patterns. The DPBA scheme must continuously adapt to these changes in real-time, which may require robust monitoring and adaptive mechanisms to handle sudden traffic spikes and ensure consistent performance.

• 3. Interference and Signal Quality: Wireless networks are prone to interference and signal quality variations. The DPBA scheme must incorporate mechanisms to account for these factors, ensuring that critical MTC applications maintain reliable communication even in challenging conditions.

• 4. Integration with Existing Infrastructure: Integrating the DPBA scheme with existing network infrastructure and protocols may pose challenges. Compatibility with various hardware and software components must be ensured, necessitating extensive testing and validation.

• 5. Security and Privacy: With the increased prioritization of certain data streams, ensuring the security and privacy of sensitive information becomes paramount. The DPBA scheme must incorporate robust security measures to protect against potential threats and breaches.

4.3.3 Future work

Future research can focus on addressing the identified challenges, particularly in enhancing the scalability and adaptability of the DPBA scheme. Additionally, exploring the integration of machine learning techniques to predict traffic patterns and optimize resource allocation dynamically can further improve the efficiency and effectiveness of the DPBA scheme. Real-world deployments and extensive field trials will be crucial in validating the scheme’s performance and identifying areas for further improvement.

The DPBA scheme represents a significant advancement in dynamic resource management for LTE networks, effectively addressing the challenges of MTC and HTC coexistence. Its ability to adapt to varying traffic demands and prioritize critical applications makes it a valuable solution for diverse IoT and communication scenarios. By exploring the potential impacts and addressing the practical challenges, this research paves the way for more efficient and robust wireless communication networks in the future.

5 Conclusion

In this research, we have presented a comprehensive investigation into the efficacy of the DPBA scheme in the context of coexisting MTC and HTC within LTE networks. Our analysis and evaluations shed light on the scheme’s performance across various scenarios, emphasizing its potential to enhance resource utilization, mitigate latency, and improve overall throughput while ensuring fairness and reliable data delivery.

We introduced the DPBA scheme as a dynamic allocation mechanism that optimally manages the allocation of resources to MTC and HTC applications. By assigning dynamic priorities to MTC applications based on HTC traffic fluctuations, DPBA achieves efficient bandwidth distribution, maintaining the desired level of service quality for both MTC and HTC services. We have derived and formulated the optimization problem for the DPBA scheme, facilitating the maximization of overall throughput while adhering to latency and fairness constraints.

Through extensive simulations conducted in MATLAB, we evaluated the DPBA scheme’s performance against existing schemes such as Proportional Fairness (PF), Static Priority Scheduling (SPS), and Machine Learning-based Scheme (MLS). Our findings highlighted DPBA’s superiority in terms of throughput, latency, fairness, and packet loss rates. The DPBA scheme consistently demonstrated better performance across these metrics, showcasing its adaptability to varying traffic conditions and its ability to deliver reliable and efficient communication services.

The DPBA scheme’s favorable performance underscores its potential to significantly contribute to the advancement of LTE networks, particularly in scenarios involving coexisting MTC and HTC applications. The scheme’s dynamic nature, prioritization mechanism, and robust performance make it a promising solution for addressing the challenges of resource allocation in heterogeneous communication environments. Future research could explore the integration of machine learning techniques to further enhance the DPBA scheme’s adaptability and efficiency.

In conclusion, the DPBA scheme emerges as a strong candidate for efficient bandwidth allocation in LTE networks, fostering improved quality of service for both MTC and HTC applications. Its adaptability, fairness, and performance gains mark a significant step toward addressing the complexities of modern wireless communication networks, positioning DPBA as a valuable tool for optimizing resource utilization and ensuring reliable communication in diverse traffic scenarios.