5G-NR PHY Software Interface for Scalable Multi-Cell Processing

The patent segments the PHY signal processing into multiple independent pipelines (e.g., downlink pipelines, uplink pipelines, multi-user MIMO pipelines) that can be executed in parallel on different PPUs. This segmentation allows the system to handle multiple users and cells simultaneously without requiring all processing resources to be allocated at once, thereby improving productivity while managing memory resources more efficiently through parallel rather than sequential allocation.

The patent introduces a new dimension of parallel processing by utilizing multiple PPUs to execute signal processing pipelines simultaneously. Instead of increasing memory resources linearly with additional users, the system transitions from single-threaded sequential processing to multi-threaded parallel processing across multiple processing units, effectively adding a temporal and spatial dimension to resource utilization that decouples memory requirements from user/cell count.

The patent divides the signal processing workload into segmented pipelines that can be executed concurrently on multiple PPUs. By segmenting processing tasks into independent units (e.g., separate downlink and uplink pipelines, separate multi-user MIMO pipelines), the system reduces the total execution time through parallel processing, thereby improving productivity without proportionally increasing time resource consumption.

The patent enables continuous processing by maintaining multiple active pipelines that operate simultaneously on different data sets. While one pipeline processes data for a group of users, another pipeline can concurrently process data for a different group, ensuring that processing action is continuous and uninterrupted. This continuity reduces idle time and improves overall productivity without requiring additional time resources.

The patent designs the PPU architecture and software framework to be universal and multi-functional, where the same PPU can execute different types of signal processing pipelines (downlink, uplink, multi-user MIMO) depending on configuration. This universality allows the system to achieve high throughput with fewer PPUs rather than requiring specialized hardware for each function, thereby reducing device complexity while maintaining productivity improvements.

The patent introduces a software framework as an intermediary layer between the PPUs and the signal processing tasks. This framework handles the complexity of pipeline management, resource allocation, and task scheduling, allowing the PPUs themselves to remain relatively simple execution units. The intermediary software absorbs much of the system complexity, enabling high throughput processing without proportionally increasing hardware device complexity.