5.77 → 17.99 dB
CONTROLLED SIMULATION · SUPPLIED PROJECT RESULTSA clearer signal.
A clearer signal.
A measurable difference.
The supplied before/after summary shows higher mean SINR and throughput, with lower mean latency and BLER. These are simulation results, not independently verified lab or field performance.
EXPLORE THE RESULTSClick any metric to switch its view
96.00 → 198.08 Mbps
151.99 → 9.72 ms
45.59% → 6.05%
SINR
Mean · dB
Throughput
Mean · Mbps
Latency
Mean · ms
BLER
Mean · %
Each metric uses its own scale. SINR bars show reported dB means; relative improvement is not inferred from bar length.
All six KPIs. Full context.
Inspect means, medians and ranges, or download the original summary for technical review.
Download KPI summary (CSV) ↓| KPI | Unit | Mean (Before) | Mean (After) | Δ Mean | Improvement % (supplied) | Median (Before) | Median (After) | Min (Before) | Min (After) | Max (Before) | Max (After) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| SINR | dB | 5.7721 | 17.9876 | 12.2155 | 211.6296 | 3.6893 | 16.9341 | -7.9954 | -0.6143 | 129.0681 | 145.1075 |
| Throughput | Mbps | 95.9974 | 198.0841 | 102.0867 | 106.3432 | 24.3485 | 93.0424 | 2.9743 | 12.6219 | 2401.0277 | 2699.4051 |
| Latency | ms | 151.9868 | 9.7195 | -142.2673 | 93.6050 | 9.4373 | 9.0000 | 9.0000 | 9.0000 | 1008.0000 | 85.4944 |
| BLER | ratio | 0.4559 | 0.0605 | -0.3953 | 86.7181 | 0.3043 | 0.0000 | 0.0000 | 0.0000 | 0.9990 | 0.9871 |
| Diversity | dB | 11.7927 | 24.0082 | 12.2155 | 103.5850 | 9.7099 | 22.9547 | -1.9748 | 5.4063 | 135.0887 | 151.1281 |
| Multiplexing | rank | 1.7578 | 1.8496 | 0.0918 | 5.2222 | 1.0000 | 1.0000 | 1.0000 | 1.0000 | 4.0000 | 4.0000 |
HOW TO READ THESE RESULTSPromising gains.
Promising gains.
Clearly defined boundaries.
- SINR & diversity: headline gains are expressed as +12.2155 dB. The supplied percentage changes of dB values are retained in the downloadable data for traceability, not used as linear power gains or coverage claims.
- Latency: the supplied table labels this metric “Latency.” The project kickoff deck identifies this as modeled latency in the simulation pipeline; it is not a measured end-to-end application delay.
- BLER: values are ratios. The mean changes from 0.4559 to 0.0605 (45.59% to 6.05%). A BLER reduction alone does not establish the same percentage reduction in retransmissions or energy consumption.
- Precision: headline relative changes follow the supplied improvement column. Small discrepancies with recalculation from displayed means may reflect rounding. The supplied BLER Δ mean is −0.3953; subtracting the displayed means gives −0.3954.
- Distribution: averages do not describe every link. The supplied maxima of 129.0681 / 145.1075 dB SINR and 135.0887 / 151.1281 dB diversity warrant a model and outlier review before engineering conclusions are drawn.
VALIDATION CONTEXT
What the next report needs.
The project kickoff deck describes six base stations, 1,024 user locations on a 32 × 32 grid, 3.5 GHz / 20 MHz, 2 dB Gaussian measurement uncertainty, a three-layer CNN, rank 1–4 adaptation and an eight-element ULA. Repeated-run counts, uncertainty and inference overhead still need documentation. Source: KashefTel project kickoff presentation supplied by the project.
Reliability hypothesis
Lower simulated BLER motivates testing retransmissions and delivery reliability in matched radio conditions.
Capacity hypothesis
Higher mean throughput motivates testing useful goodput under realistic load. It does not establish doubled cell capacity.
Business hypothesis
Coverage, site count, battery life, CAPEX and customer revenue need separate models or measurements. They are not quantified by this table.