Uplink 256QAM Activation over LTE and 5G NR

Uplink 256QAM Activation over LTE and 5G NR

Abdul Quader Syed (1), Mohammed Babar Ahmed (2), Hasan Omair Mohammed (3)

Nokia Solution & Networks

Abstract – This paper presents the results of a eld Proof-of-Concept (PoC) trial evaluating the activation of Uplink (UL) 256

Quadrature Amplitude Modulation (256QAM) across 64 base-station sites in the Kingdom of Saudi Arabia. The feature was simultaneously enabled on September 29, 2025, covering all LTE (4G) and 5G New Radio (NR) cells across eight postal-code zones in Makkah, Madinah, Yanbu, Turbah, Abha, and Rabigh. Performance was assessed using two independent measurement methodologies: Operations Support System (OSS) counter-based Key Performance Indicator (KPI) analysis and crowdsourced Ookla Speedtest Intelligence data. Results indicate a 15% in-crease in peak daily uplink speed and a 20% improvement in Ookla Speed Score for devices supporting advanced modem chipsets. OSS measurements also reveal gains of 5% in the maximum uplink throughput of 5G, 4.5% in the average uplink throughput of 5G, 10% in the maximum uplink throughput of LTE, and 7.5% in the average uplink throughput of LTE. Following activation, UL 256QAM accounted for 14% of 5G uplink transmissions and 22% of LTE uplink transmissions. The ndings demonstrate that UL 256QAM delivers signicant uplink capacity and throughput enhancements while requiring no additional spectrum or hardware investment, making it an effective and spectrum-efcient software-driven upgrade for existing 4G and 5G networks.

Index TermsUL 256QAM, 5G NR, LTE, Uplink Throughput, Network Optimization, OSS KPI Analysis, Speedtest Intelligence.

1. INTRODUCTION

   The continuous growth of mobile broadband trafc and the increasing demand for uplink-intensive applications such as cloud services, video sharing, live streaming, and industrial Internet of Things (IoT) have intensied the need for more efcient utilization of radio spectrum resources. Higher-order modulation schemes represent a key mechanism for improving spectral efciency and increasing network capacity without requiring additional spectrum allocation.

   Among these schemes, 256 Quadrature Amplitude Mod-ulation (256QAM) has been widely adopted for downlink transmissions in both Long-Term Evolution (LTE) and Fifth-Generation New Radio (5G NR) systems. By encoding 8 bits per symbol, 256QAM provides a theoretical spectral efciency improvement of approximately 33% over 64QAM, which carries 6 bits per symbol. However, uplink deployment of 256QAM has historically been limited due to stringent signal-to-interference-plus-noise ratio (SINR) requirements, user equipment (UE) transmit power constraints, and the limited availability of compatible device chipsets.

   Recent advancements in modem technology and radio-frequency front-end design have signicantly improved UE uplink transmission capabilities, enabling broader support

   for UL 256QAM in commercially available devices. Conse-quently, network operators can now leverage UL 256QAM as a software-based enhancement to increase uplink throughput and capacity while minimizing infrastructure investment.

   Despite its growing availability, limited eld-scale studies have quantied the real-world performance impact of UL 256QAM in operational LTE and 5G networks. Most ex-isting investigations focus on theoretical analysis, link-level simulations, or laboratory testing, leaving a gap in large-scale deployment validation under live network conditions. This mo-tivates the need for comprehensive eld evaluations using both network-side and user-experience-based performance metrics. To address this gap, we conducted a Proof-of-Concept (PoC) trial involving the activation of UL 256QAM across 64 LTE and 5G sites distributed over eight postal-code regions in the Kingdom of Saudi Arabia. The trial assessed performance using two complementary measurement approaches: Opera-tions Support System (OSS) Key Performance Indicator (KPI) analysis and crowdsourced Ookla Speedtest Intelligence data. The combination of these methodologies enables evaluation from both network performance and end-user experience per-

   spectives.

   The main contributions of this work are summarized as follows:

   • Development of a dual-methodology evaluation frame-work combining OSS counter analysis and crowdsourced speed-test measurements for uplink performance assess-ment.

   • Large-scale empirical evaluation of UL 256QAM across 64 commercial LTE and 5G sites spanning eight geo-graphical regions in Saudi Arabia.

   • Quantication of throughput improvements in both LTE and 5G NR networks following UL 256QAM activation.

   • Analysis of modulation utilization patterns across radio technologies, highlighting practical adoption levels and deployment-dependent behavior.

   • Provision of deployment recommendations demonstrating the viability of UL 256QAM as a spectrum-efcient software upgrade for existing mobile networks.

2. BACKGROUND AND RELATED WORK

   1. 256QAM in Cellular Standards

      The Third Generation Partnership Project (3GPP) intro-duced Uplink (UL) 256 Quadrature Amplitude Modulation (256QAM) in LTE Release 13 and subsequently incorporated support within the 5G New Radio (NR) framework in Release

      15. The feature is available only for user equipment (UE) cat-egories and capability classes that support higher-order uplink modulation and is subject to network scheduler conguration and radio channel conditions. By increasing the modulation order from 64QAM to 256QAM, the number of transmitted bits per symbol rises from six to eight, enabling a theoretical spectral efciency improvement of approximately 33%.

      In practical deployments, the selection of the uplink mod-ulation scheme is dynamically determined by the eNB/gNB scheduler based on radio link quality measurements reported by the UE, including Channel Quality Indicator (CQI), Rank Indicator (RI), and other link adaptation parameters. In LTE systems, UL 256QAM is associated with the highest Modula-tion and Coding Scheme (MCS) levels dened in the standard, while in 5G NR it is supported through the higher-order MCS tables specied by 3GPP. Consequently, UL 256QAM is typically utilized only when channel conditions provide suf-ciently high Signal-to-Interference-plus-Noise Ratio (SINR) and adequate uplink power margins to maintain acceptable block error rates (BLER).

      The achievable throughput gain from UL 256QAM depends on several factors, including radio propagation conditions, interference environment, scheduler implementation, device capability, and trafc loading. Therefore, eld-based perfor-mance validation in commercial networks is essential to quan-tify the practical benets of UL 256QAM beyond theoretical spectral-efciency improvements.

   2. Prior Field Studies

   Although the benets of higher-order modulation schemes have been extensively investigated for downlink transmissions, published studies examining the real-world performance of Uplink (UL) 256QAM remain relatively limited. Most ex-isting research has focused on theoretical analysis, link-level simulations, or small-scale laboratory evaluations. Simulation-based studies have projected uplink spectral efciency im-provements ranging from 15% to 30% under favorable ra-dio conditions characterized by high Signal-to-Interference-plus-Noise Ratio (SINR) levels exceeding 25 dB. Similarly, a limitd commercial-network trial reported approximately 12% improvement in LTE uplink throughput following UL 256QAM activation across a small set of urban deployment sites.

   Despite these encouraging results, there remains a lack of large-scale eld validation covering heterogeneous geograph-ical environments, multiple radio access technologies (RATs), and diverse user populations. Furthermore, few studies have combined network-side performance indicators with end-user experience metrics to provide a comprehensive assessment of UL 256QAM deployment benets.

   To address these limitations, this work presents a large-scale commercial-network evaluation of UL 256QAM conducted across 64 LTE and 5G sites spanning eight postal-code regions in the Kingdom of Saudi Arabia. The study integrates Opera-tions Support System (OSS) KPI analysis with crowdsourced Ookla Speedtest Intelligence measurements, enabling simulta-neous assessment of network performance and user-perceived throughput.

3. TRIAL DESIGN AND METHODOLOGY

   1. Site Selection and Deployment

      The trial comprised 64 radio sites selected to represent a cross-section of the Nokia-managed network in Saudi Arabia, covering urban, semi-urban, and suburban morphologies. Sites were activated simultaneously on 29 September 2025 with UL 256QAM enabled on all co-located 4G and 5G cells. No other conguration changes were introduced during the measurement window. Table I summarizes the geographic distribution.

      TABLE I

      GEOGRAPHIC DISTRIBUTION OF TRIAL SITES

      District City Postal Code No. of Sites

      MAKKAH Makkah 2XXXX 16

      MAKKAH Makkah 2XXXX 3

      MADINAH Madinah 4XXXX 20

      MADINAH Rabigh 2XXXX 4

      YANBU Yanbu 4XXXX 7

      TAIF Turbah 2XXXX 3

      ASSIR Abha 6XXXX 4

      ASSIR Abha 6XXXX 6

      Total 8 Zones 64

   2. Measurement Methodology

      Performance was assessed using two independent method-ologies to ensure robustness of ndings:

      • OSS KPI Analysis: Network-side counters extracted from the Operations Support System were used to evaluate peak and average UL throughput, modulation scheme uti-lization rates, and uplink resource slot utilization before and after the feature activation date.

      • Ookla Speedtest Intelligence: Third-party crowdsourced speed test data from Ookla was analyzed for UL speed trends (Mbps) and the composite Ookla Speed Score, segmented by access technology (4G/5G) and chipset generation. Weekly and daily peak values as well as average Speed Scores were computed for the pre- and post-activation periods.

   The pre-activation baseline spans 1 July 28 September 2025 (13 weeks). The post-activation assessment window is 29 September 12 October 2025 (2 weeks). Statistical comparisons use the mean of daily peak values for the respective periods.

4. OOKLA PERFORMANCE ANALYSIS

   1. UL Speed Improvement

      Table II presents the Ookla UL speed statistics pre- and post-activation. Modern chipset devices exhibited the largest absolute and relative gains, consistent with the expectation that high-end UEs are more likely to possess and leverage UL 256QAM encoding capability.Table II summarizes the Ookla UL speed.

      TABLE II

      Ookla UL Speed Results (Mbps)

      Metric Pre Post %

      Modern Chipset Wkly. Peak 21.2 24.3 +14.6%

      Modern Chipset Daily Peak	33.5	38.6	+15.0%
      5G Weekly Peak	22.5	25.7	+13.2%
      5G Daily Peak	33.9	38.0	+12.0%
      4G Weekly Peak	12.0	13.1	+9.1%

      4G Daily Peak 29.8 32.0 +7.0%

   2. Speed Score Analysis

   The Ookla Speed Score is a composite metric that weighs both download and upload speeds using a logarithmic scale, providing a more holistic representation of user-experienced network performance. Post-activation Speed Scores demon-strated signicant uplift, particularly for 5G and modern chipset users.Table III summarizes the Ookla speed score.

   TABLE III

   Ookla Speed Score (Average)

   Segment Pre Post %

   Modern Chipset 169 203 +20%

   5G 200 232 +16%

   4G 33 36 +9%

   The +20% improvement in Speed Score for modern chipset devices is particularly noteworthy, as it reects the com-pounded benet of 256QAM UL adoption among devices most capable of exploiting the higher modulation order. The more modest 4G uplift (+9%) aligns with expectations given the typically lower SINR conditions experienced by LTE UL users relative to their 5G counterparts.

5. 5G NR UL 256QAM OSS ANALYSIS

   1. Throughput and Utilization

      5G UL throughput, measured without EN-DC contribution to isolate NR scheduler behaviour, increased by 5% at the cell peak level and by 4.5% on average following 256QAM activation (Table IV). Concurrently, UL PUSCH data slot utilization decreased by approximately 1 percentage point, indicating that the same or greater data volume is being carried with fewer allocated resource units a direct manifestation of improved spectral efciency.

      TABLE IV

      5G NR OSS KPI SUMMARY

      5GKPI Max UL Throughput gain	Value +5.0%
      Avg. UL Throughput gain	+4.5%
      UL Slot Utilization change	1.0%
      256QAM avg. UL share	14%
      256QAM UL range (by band)	13%37%

      QPSK share reduction 3.7%

   2. Modulation Scheme Distribution

   Prior to activation, NR UL transmissions were dominated by QPSK (50.8%) and 64QAM (16.4%), with no 256QAM usage.

   Post-activation, 256QAM accounts for 14.1% of UL transmis-sions on average. The QPSK share fell to 28.8% and 64QAM declined to 15.6%, conrming active scheduler exploitation of the higher modulation order when CQI conditions permit [6]. Band-level 256QAM utilization varies signicantly: NR700 cluster records 37%, NR3500 13%, and NR2300 10% [7]. The NR700 result is attributable to wider coverage areas and generally higher UE receive power, both of which elevate

   reported UL CQI.

6. LTE UL 256QAM OSS Analysis

   LTE UL throughput was evaluated with Carrier Aggregation (CA) active. Peak throughput (Max PDCP) improved by 10%, and average user throughput increased by 7.5% (Table V). The larger absolute gains in LTE relative to 5G are consistent with the fact that LTE 64QAM was the pre-activation ceiling in most conditions, while 5G NR already supported per-carrier aggregation and wider bandwidths that partially mitigate the modulation order bottleneck.

   TABLE V

   LTE OSS KPI SUMMARY

   4GKPI Value Max UL Throughput gain (with CA) +10.0% Avg. UL Throughput gain +7.5%

   256QAM avg. UL share 22%

   256QAM UL range (by band) 11%34%

7. MODULATION UTILIZATION BY BAND

   LTE achieves a higher aggregate 256QAM utilization (22%) than 5G (14%), which is attributed to the larger installed base of 256QAM-capable LTE devices in the trial footprint and the relatively mature LTE scheduler parameter tuning. Band-level results show LTE2300 at 34%, LTE1800 at 28%, LTE2100 at 25%, LTE900 at 15%, and LTE00 at 11%, with the higher-frequency bands beneting from denser deployments and lower inter-cell interference in these specic areas [8].

8. DISCUSSION

   1. Consistency of Dual Measurement Pipelines

      Both OSS KPI and Ookla data independently conrm positive performance impacts, providing high condence in the reliability of the results. The Ookla Speed Score gain (+20% modern chipset) exceeds the OSS throughput gain (+5% 5G,

      +10% LTE) because the Speed Score is a peak-of-distribution metric, while OSS throughput represents a cell-average and is diluted by users in poor channel conditions who cannot utilise 256QAM.

   2. Efciency vs. Throughput Trade-off

      The 5G UL slot utilization reduction (-1%) alongside throughput improvement (+4.5%) demonstrates that UL 256QAM activation delivers spectral efciency gains without requiring additional frequency resources. This is particularly signicant given the increasing scarcity of licensed spectrum.

   3. Device Ecosystem Dependency

      The gap between modern-chipset and all-device improve-ments (e.g., +15% vs. lower averages for UL speed) un-derscores that the full benet of UL 256QAM is device-ecosystem-dependent. As the installed base of capable UEs grows, network-level gains are expected to increase progres-sively.

   4. Limitations

      The trial was conducted across a xed set of 64 sites, and results may not be fully representative of the broader network. Seasonal variations in trafc patterns, geographic-specic propagation characteristics, and the evolving device ecosystem may inuence long-term performance metrics.

   5. Key Observations

   LTE demonstrated higher 256QAM UL utilization (22%) compared to 5G (14%), attributed to the larger base of 256QAM-capable LTE devices in the trial areas. The largest throughput gains were observed in LTE (+10% max through-put), where 256QAM provides a more signicant relative improvement over the legacy 64QAM ceiling. Modern chipset segmentation in Ookla data reveals that the benets of 256QAM are disproportionately experienced by users with high-end devices, underscoring the device ecosystem depen-dency of UL modulation order support. The reduction in QPSK share (3.7% for 5G) and concurrent increase in 256QAM usage conrms that the scheduler is actively leveraging the higher modulation order for eligible users rather than simply enabling the feature without utilization. The decrease in 5G UL slot utilization (-1%) alongside throughput gains indicates improved spectral efciency a key objective of advanced modulation deployment.

9. CONCLUSION

   This paper presented UL 256QAM activation represents a cost-effective, software-driven network enhancement that delivers measurable improvements in both network KPIs and end-user experience metrics. The key conclusions are as fol-lows:

   • Ookla UL speed increased by up to +15% for modern chipset devices, with Speed Score improving by +20%.

   • OSS KPIs conrm +5%, +4.5% (max/avg) 5G UL throughput and +10% +7.5% (max/avg) LTE UL through-put improvements.

   • 256QAM UL utilization reached 14% (5G) and 22% (LTE), with signicant band-level variation reecting propagation and deployment differences.

   • A 5G UL slot utilization reduction of 1% demonstrates improved spectral efciency, enabling capacity gain with-out additional spectrum.

   Based on these results, network-wide deployment of UL 256QAM is recommended. Future work should include a 90-day longitudinal study, quality-of-service impact assessment, and analysis of UL 256QAM interaction with uplink MIMO and carrier aggregation features.

10. ABBREVIATIONS

The following abbreviations are used in this manuscript: 256QAM 256 Quadrature Amplitude Modulation

CA Carrier Aggregation CQI Channel Quality Indicator DL Downlink

EN-DC E-UTRAN New Radio Dual Connectivity KPI Key Performance Indicator

LTE Long-Term Evolution MCS Modulation and Coding Scheme NR New Radio

OSS Operations Support System PDCP Packet Data Convergence Protocol PoC Proof of Concept

PUSCH Physical Uplink Shared Channel QPSK Quadrature Phase Shift Keying RAT Radio Access Technology

SINR Signal-to-Interference-plus-Noise Ratio UE User Equipment

UL Uplink

PR Physical Resource Block

TDD Time Division Duplex FDD Frequency Division Duplex QoS Quality of Service

Mbps Megabits per Second OOKLA Speedtest Intelligence Platform

REFERENCES

1. 3GPP. Physical layer procedures (LTE). TS 36.213, Release 16; 3rd Generation Partnership Project: Sophia Antipolis, France, 2020.

2. 3GPP. Physical layer procedures for data (NR). TS 38.214, Release 16; 3rd Generation Partnership Project: Sophia Antipolis, France, 2020.

3. Dahlman, E.; Parkvall, S.; Sko¨ld, J. 5G NR: The Next Generation Wireless Access Technology, 2nd ed.; Academic Press: London, UK, 2021.

4. 3GPP TS 38.331, NR; Radio Resource Control (RRC) Protocol Spec-ication, Release 17, 2023.

5. 3GPP TR 21.916, NR Inter-band Carrier Aggregation and Dual Con-nectivity, Release 16, 2022

6. Mohammed Yahiya Pasha Gulam Performance Analysis of 5G N71 Deployment in Live Network Environments, IJERT Volume 15, Issue 04 , April 2026

7. Mohammed Yahiya Pasha Gulam Performance Evaluation of N40 Integration Using AZNA in 5G Networks, IJERT Volume 15, Issue 04 , April 2026

8. Mohammed Yahiya Pasha Gulam 2L-4H Beams Antenna, IJERT Volume 15, Issue 05 , May 2026

9. Report M.2370-0; International Telecommunication Union: Geneva, Switzerland, 2015.

10. Tullberg, H.; Popovski, P.; Li, Z.; Uusitalo, M.A. The METIS 5G System Concept: Meeting the 5G Requirements. IEEE Commun. Mag. 2016, 54, 132139

11. Ookla. Speedtest Intelligence Methodology White Paper; Ookla LLC: Seattle, WA, USA, 2024. Available online: https://www.ookla.com/articles/speedtest-intelligence (accessed 12

October 2025).