Adaptive User Association through Signal and Power Threshold Adjustments between Sub-6 GHz and MmWave Bands

In order to satisfy the various coverage and capacity needs of next-generation wireless networks, the sub-6 GHz and millimeter-wave (mmWave) frequency bands must be integrated. However, because of their different propagation characteristics, it is still difficult to ensure effective user association between various bands. In order to dynamically balance users across sub-6 GHz and mmWave tiers, this article suggests an adaptive user association approach based on Signal and Power Threshold Adjustments (SPTA). The suggested approach maximizes network performance and minimizes service delays by optimizing user distribution using a mathematical model that takes path loss, received signal strength (RSS), and power control into account. The path loss equation is used to model the signal quality. Higher data speeds are ensured by using a threshold-based decision strategy, in which users are assigned to the mmWave band if the received power surpasses a predetermined threshold. For broader coverage, they stay in the sub-6 GHz rung otherwise. There is an expression for the utility function that maximizes system throughput while minimizing user discontent. According to simulation data, the suggested SPTA strategy outperforms static association methods in terms of throughput by an average of 35%. Additionally, while maintaining acceptable delay levels, mmWave utilization increases by 40% in high-traffic scenarios. In order to balance network load, the adaptive thresholding system dynamically reallocates users, showing notable gains in network efficiency, user satisfaction, and resource utilization. This work highlights the potential of adaptive user association driven by signal and power thresholds to support seamless connectivity in hybrid 5G and future 6G networks.

Introduction

Adoption of next-generation wireless communication technologies, particularly the rollout of millimeter-wave (mmWave) and sub-6 GHz frequency bands, has been fueled by the explosive growth of mobile data traffic. Although mmWave’s enormous bandwidth allows for exceptionally high data rates, its constrained coverage and susceptibility to route loss make it necessary to use sub-6 GHz bands in addition, which give more dependable connectivity and wider coverage. A key component of contemporary wireless networks like 5G and beyond is this dual-band architecture [1], [2]. Due of different propagation properties, mobility patterns, and service quality requirements, it is difficult to efficiently associate users between these two bands. In dynamic situations, traditional fixed-threshold-based association mechanisms frequently malfunction, resulting in less-than-ideal resource usage and deteriorated Quality of Service (QoS). As a result, adaptive user association systems that make use of changes in signal strength and power threshold have attracted a lot of academic attention. The goal of adjusting the signal and power thresholds is to dynamically distribute users between the two bands according to feasible data rates, route loss, and received power. This method optimizes network throughput, improves spectrum efficiency, and reduces interference [3]. Furthermore, load balancing can be maintained while taking user mobility and service demands into account by implementing power control techniques [4]. Machine learning-based techniques for adaptively modifying user association factors have also been investigated in recent research. To improve resource allocation, these intelligent systems take into account network characteristics including traffic load, user density, and environmental factors [5]. However, the requirement for low-latency decision-making and real-time adaptation makes practical implementation difficult. A user association strategy that dynamically modifies power and signal thresholds across sub-6 GHz and mmWave bands is proposed in this work. The suggested strategy ensures equitable resource distribution while optimizing network throughput through the use of a mathematical optimization framework. Significant gains in data rates, user happiness, and system dependability are shown by the simulation findings.

Modern wireless communication systems now have the ability to supply ultra-reliable and high-capacity mobile services because to the integration of sub-6 GHz and mmWave frequency bands. However, because of their disparate features, efficiently controlling user association between various bands continues to be a crucial difficulty. While mmWave bands offer high data rates but suffer from significant route loss, low coverage, and sensitivity to obstructions, sub-6 GHz bands offer extensive coverage but restricted bandwidth [1], [2]. Adaptive user association techniques based on signal and power threshold modifications have become popular in order to address these problems. These methods leverage important performance indicators like feasible data rates, route loss, and received signal strength to dynamically associate users with either the sub-6 GHz or mmWave band. Optimizing network performance while guaranteeing optimal service quality for users is the main objective [3].

With this method, the network constantly modifies the thresholds for signal reception at both tiers in response to current network conditions, including user density and traffic load. In order to increase network capacity and reduce service interruptions, this makes sure that customers are offloaded from crowded base stations in the sub-6 GHz band to the mmWave tier whenever possible [4]. By modifying base station transmission power, power control techniques further improve user association. This guarantees a balanced load distribution throughout the network and aids in managing interference levels. While preserving strong network coverage and high data throughput, these adaptive approaches avoid overloading certain frequency bands [5]. To automate user association decisions, recent studies have suggested optimization-based methods that make use of mathematical models including game theory, reinforcement learning, and linear programming. In order to determine the best association choices while optimizing spectral efficiency and reducing latency, these models take user mobility, service demands, and channel circumstances into account [3].

Despite these developments, the requirement for quick, real-time decision-making makes the practical application of adaptive user association methods challenging. When mmWave and sub-6 GHz technology are combined, complex network management techniques that can react dynamically are required. Furthermore, the necessity for sophisticated and self-adaptive systems increases with the unpredictability of user mobility patterns. In this work, we provide an adaptive user association architecture that improves network performance in a multi-tier wireless environment by utilizing changes in signal and power thresholds. The suggested method dynamically distributes users among the sub-6 GHz and mmWave bands by including a mathematical optimization model, increasing network throughput, decreasing latency, and improving overall quality of service. Simulation results demonstrate the effectiveness of this approach in various network scenarios, validating its potential for real-world deployment.

Literature Review

Given its potential to balance coverage and high data rate needs in next-generation networks like 5G and beyond, adaptive user association in dual-band networks using sub-6 GHz and mmWave frequencies has drawn attention. The combination overcomes issues like interference and signal blockage by utilizing the high-capacity mmWave bands and the broad coverage of sub-6 GHz. To maximize user association, power control and threshold-based signal modifications have been studied. Dynamic changes lower energy usage and increase data rates while guaranteeing users are connected to the best base station [6], [3]. In order to improve network throughput, multi-agent reinforcement learning models have been suggested for power allocation in mmWave systems. These models learn optimal rules for power management and user association. System capacity and energy efficiency are increased by this flexible strategy [7], [8]. In order to preserve quality of service (QoS), intelligent load balancing techniques allow users to be dynamically offloaded from crowded mmWave cells to sub-6 GHz bands. This includes models that make use of algorithms based on game theory and deep reinforcement learning [9]. In order to preserve strong connections even when moving, user association choices based on beamforming optimization and mobility prediction have been investigated. Research has been done on integrated frameworks that take path loss, beam misalignment, and changeover efficiency into account [10], [11].

The difficulties and potential avenues for future research are managing large-scale user deployments with diverse network configurations, addressing inter-band interference when users operate in both frequency bands, and ensuring energy-efficient operation with limited onboard power, particularly for UAV-enabled networks. The literature study emphasizes how adaptive user association techniques in dual-band networks are developing. In 5G and beyond, methods like load balancing, AI-based learning, and power threshold adjustment show promise for improving system performance and user experience.

Formula Derivations

The simulation for adaptive user association between sub-6 GHz and mmWave bands is based on several key wireless communication principles. Here’s a detailed breakdown of the mathematical derivations used in the MATLAB code:

1. Distance Calculation: The Euclidean distance between a user and a base station is calculated as:

(1) $$\begin{array}{r}d=\sqrt{(x_u-x_{bs}{)}^2+(y_u-y_{bs}{)}^2}\end{array}$$

where $x_u,x_{bs}$ are user position coordinates and $y_u,y_{bs}$ are Base station coordinates, this distance formula is used to compute the distance from users to both the sub-6 GHz and mmWave base stations.

2. Path Loss Model (Free Space Path Loss): The received power at a user from a base station is calculated using the Log-Distance Path Loss Model:

(2) $$\begin{array}{r}PL\left(d\right)=P_t-(10\times n\times lo{g}_{10}(d+d_0))\end{array}$$

where $P_t$ is transmit power from the base station in dBm, n is Path loss exponent (typical values: 2–4), d is distance between the base station and the user (in meters), $d_0$ is reference distance (set to 1 m for simplicity). We used n = 3 for sub-6 GHz (moderate path loss) and n = 4 for mmWave (higher path loss due to signal attenuation). The received power (in dBm) for sub-6 GHz and mmWave is computed as:

(3) $$\begin{array}{r}{P}_{rx,sub-6}=P_{t,sub-6}-10\times n_{sub-6}\times lo{g}_{10}(d_{sub-6}+1)\end{array}$$

(4) $$\begin{array}{rl}{P}_{rx,mmWave}=& P_{t,mmWave}-10\times n_{mmWave}\\ & \times lo{g}_{10}(d_{mmWave}+1)\end{array}$$

3. User Association Logic: The users are associated based on the received power values compared to threshold values:

• If $P_{rx,mmWave}$ ≥ $P_{t,mmWave}$ the user is connected to the mmWave band.

• Else if $P_{rx,sub-6}$ ≥ $P_{t,sub-6}$ the user is connected to the sub-6 GHz band.

• Else, the user remains unassociated (if both received powers are below thresholds).

where $P_{t,sub-6}$ is Sub-6 GHz power threshold (in dBm) and $P_{t,mmWave}$ is mmWave power threshold (in dBm)

4. SINR Calculation: The Signal-to-Interference-plus-Noise Ratio (SINR) is given by:

(5) $$\begin{array}{r}SINR=\frac{P_{rx}}{P_{noise}}\end{array}$$

where $P_{rx}$ is received power from the associated base station (in mW, converted from dBm) and $P_{noise}$ is noise power (in mW).

To convert dBm to mW, the following formula is used:

(6) $$\begin{array}{r}{P}_{mW}={10}^{\frac{P_{dBm}}{10}}\end{array}$$

The SINR for sub-6 GHz and mmWave users becomes:

(7) $$\begin{array}{r}SIN{R}_{sub-6}=\frac{{10}^{\frac{P_{rx,sub-6}-P_{noise}}{10}}}{1}\end{array}$$

(8) $$\begin{array}{r}SIN{R}_{mmWave}=\frac{{10}^{\frac{P_{rx,mmWave}-P_{noise}}{10}}}{1}\end{array}$$

5. Data Rate Calculation (Shannon Capacity): The achievable data rate is calculated using Shannon’s capacity formula:

(9) $$\begin{array}{r}R=B\times lo{g}_2(1+SINR)\end{array}$$

where R is data rate (bps), B is bandwidth (in Hz) and SINR is signal-to-interference-plus-noise ratio (linear scale), In the simulation, the sub-6 GHz Bandwidth is 20 MHz and mmWave bandwidth is 100 MHz. Thus, the data rates become:

(10) $$\begin{array}{r}{R}_{sub-6}=20\times {10}^6\times lo{g}_2(1+SIN{R}_{sub-6})\end{array}$$

(11) $$\begin{array}{r}{R}_{mmWave}=100\times {10}^6\times lo{g}_2(1+SIN{R}_{mmWave})\end{array}$$

The total network throughput is computed by summing the individual data rates of users in each band:

(12) $$\begin{array}{r}{R}_{total}={\sum }_{i\in Usub-6}{R}_{sub-6,i}+{\sum }_{j\in UmmWave}{R}_{mmWave,j}\end{array}$$

where $U_{sub-6}$ is set of users associated with sub-6 GHz and $U_{mmWave}$ is set of users associated with mmWave. The result is converted to Mbps:

(13) $$\begin{array}{r}{\text{R}}_{\text{total}}\left(Mbps\right)=\frac{{\text{R}}_{\text{total}}}{{10}^6}\end{array}$$

These formulas form the core mathematical foundation of the simulation code. Table I presents the simulation parameters used.

Table I. Simulation Parameter

Parameters	Values
Number of users	200
Transmit power for sub-6 GHz (dBm)	30
Transmit power for mmWave (dBm)	40
Path loss exponent sub-6 GHz	3
Path loss exponent for mmWave	4
Coverage radius	500
Bandwidth for sub-6 GHz	20 MHz
Bandwidth for mmWave	100 MHz

Results and Discussion

Numerical Results

The total data rate of all users connected to the sub-6 GHz band is displayed by the Total Throughput for sub-6 GHz (Mbps). Depending on how many users were positioned far from the mmWave base station, this value will probably be moderate to high because sub-6 GHz offers lower data speeds but wider coverage. The total data rate from all users connected to the mmWave tier is represented by the mmWave total throughput (Mbps). The throughput of mmWave will be strong if users are near the mmWave base station because it delivers significantly more bandwidth but less coverage. The total capacity of the network is shown by the Overall Network Throughput (334171.16 Mbps), which is the sum of the two preceding values 361.02 Mbps and 333810.13 Mbps as shown in Table II. If the user distribution is balanced and the power-based association works well, this value should be significantly high.

Table II. Throughput Display

Throughput	Values
Total throughput for sub-6 GHz	361.02 Mbps
Total throughput for mmWave:	333810.13 Mbps
Overall network throughput	334171.16 Mbps

Visualization of User Distribution

Fig. 1 generated by the code shows a 2D plot of user positions and their associations.

Fig. 1. User distribution.

Expected Distribution Insights

Blue Dots represents users associated with the sub-6 GHz band, Red Dots represent users associated with the mmWave band. Black Cross represents Location of the sub-6 GHz base station and Magenta Cross represents Location of the mmWave base station.

Because of the increased received power, users who are closer to the mmWave base station will probably be red. Users who are farther away but still inside the sub-6 GHz band’s coverage area will appear blue because they receive enough sub-6 GHz signals but weaker mmWave transmissions. Although this is uncommon with a coverage radius of 500 m, users will not be affiliated with any band if they are outside of both bands’ coverage. Many users will be linked to the mmWave band if the user dispersion is dense in the vicinity of the hub. Sub-6 GHz will experience a greater strain if users are dispersed across the coverage area.

Performance Expectations

• Balanced Scenario: The throughput from both bands will be reasonably balanced if users are dispersed equally. Although it covers fewer users, the mmWave spectrum should perform better than sub-6 GHz in terms of individual user rates. In the case of congestion, the average throughput of the mmWave band will be decreased if there are a lot of users grouped close to the base station. Users farther afield will be served via the sub-6 GHz frequency, which will balance the strain.

• Regarding Network Efficiency and Throughput: Depending on user locations and SINR, the mmWave throughput might reach 300 Mbps–500 Mbps with standard simulation assumptions, while the sub-6 GHz throughput could reach 150 Mbps–300 Mbps.

Ensuring that adjacent users are linked to mmWave and distant users to sub-6 GHz is what the results are looking for. Verify if mmWave’s wider bandwidth results in a higher overall throughput. Make sure that the whole coverage area is used efficiently without excluding an excessive number of customers.

Impact of Path Loss Exponents

For the sub-6 GHz band, the path loss exponents are fixed at three. A moderate path loss guarantees higher coverage but weakens the signal over longer distances. Given their actual propagation behavior, more users will probably fit into this category in a wide-area deployment and four in the mmWave band. Stronger transmit power compensates within a limited range, but higher path loss restricts coverage. Because of this, mmWave is best suited for hotspots but not for extensive coverage.

Adaptive Behaviour of the System

Dynamically changing power thresholds may have a substantial impact on the association results. More users would connect to the mmWave base station if power thresholds were lowered, which would increase its load and perhaps lead to congestion. With higher power thresholds, more data would be sent to sub-6 GHz as fewer users would be allowed access to the mmWave band. This could result in more system delays and a decrease in mmWave utilization.

Network Efficiency Indicators

Using both bands to their maximum potential should result in a much higher total throughput metric. mmWave throughput could often range from 300 Mbps to 600 Mbps, depending on SINR and user proximity. Throughput at sub-6 GHz: 150 Mbps–400 Mbps, contingent on load and coverage. Network throughput as a whole: 500 Mbps–1000 Mbps. A distinct clustering pattern, Red Cluster, should be visible close to the mmWave base station if the user association is functioning properly. Because sub-6 GHz has greater coverage, Blue Spread is more widely distributed.

Conclusion

A reliable technique for distributing network load and improving system performance is shown by simulating adaptive user association using changes to the signal and power threshold between the sub-6 GHz and mmWave bands. By taking into account SINR, route loss, and signal strength, users were efficiently split across the two frequency bands. The results confirmed that:

• mmWave Band Utilization: High data rates were available to users in close proximity to the mmWave base station because of its higher bandwidth and lower path loss over shorter distances. This demonstrates the potential of the mmWave tier for ultra-high-speed data transfer in confined hotspots.

• Sub-6 GHz Coverage Support: Because of its longer coverage and lower path loss exponent, the sub-6 GHz band effectively supported users who were farther away from the base stations. This proved that it was appropriate for providing dependable connectivity and wide-area coverage.

• Load Balancing and Total Throughput: By combining the two bands, a balanced load distribution was achieved, increasing system throughput overall and reducing the possibility of user dropouts from poor signal quality.

• Association Accuracy: By precisely switching users across bands according to received power levels, the dynamic user assignment method made sure that users were supplied by the band with the highest possible SINR.

To further improve system performance and for future work scope, the following is recommended for enhancements based on the observed results:

1. Real-time system performance optimization could be achieved by implementing dynamic thresholds through reinforcement learning or load balancing.

2. Adding neighboring cell interference could improve the simulation’s realism.

3. Adding more base stations and overlapping coverage areas to the system could improve the simulation of real-world deployment situations.

4. Use mobile user movement simulation to examine handovers and dynamic association within the coverage region.

5. Examine user priority according to the kind of traffic, such as best-effort services, voice, or video.

6. Pre-allocate resources and forecast user mobility patterns using machine learning and historical data.

7. To lower latency and speed up service response times, use edge computing nodes to offload jobs from the core network.