Parallel Clone Processes for Rapid Initial Heading Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing GPS-based navigation systems for pedestrians take too long to determine initial heading, often leading users to travel in the wrong direction before receiving accurate directions, due to power constraints and the stochastic nature of particle filters, which are unsuitable for mobile devices.
Innovation Solution
The system spawns multiple clone processes with identical locations and speeds but unique directions, using Kalman filters to calculate positions and terminate processes with mismatched directions, thereby quickly determining the initial heading by converging on the actual direction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS-based systems use particle filters to determine initial heading, then measurement precision is improved, but loss of time increases significantly (3-7 minutes)
Solution Approach 1:
The patent segments the initial heading determination process into multiple parallel clone processes, each exploring different directional hypotheses simultaneously. Instead of using a single sequential particle filter that takes 3-7 minutes, the system divides the computational task into parallel segments that can be evaluated concurrently, dramatically reducing the time to determine initial heading while maintaining measurement precision.
Solution Approach 2:
The patent applies preliminary action by pre-spawning multiple clone processes with predetermined directional hypotheses before actual navigation begins. These clone processes prepare multiple potential heading solutions in advance, allowing the system to quickly select the correct heading once motion data is available, rather than computing the heading from scratch when needed.
2Measurement precision
If GPS-based systems wait for user movement to collect data for initial heading determination, then measurement precision improves, but loss of time increases as users may travel in wrong direction
Solution Approach 1:
The system performs preliminary actions by spawning multiple clone processes with different directional hypotheses before the user actually moves. These processes prepare potential heading solutions in advance, so when the user begins moving, the system can quickly compare actual motion data against the pre-computed hypotheses and provide the correct direction immediately, preventing users from traveling in wrong directions.
Solution Approach 2:
The patent applies dynamics by making the clone processes adaptive and dynamic rather than static. The clone processes continuously update their hypotheses based on incoming motion data from accelerometers and other sensors, allowing the system to dynamically adjust and confirm the correct heading as the user begins to move, rather than waiting for a fixed amount of data collection.
3Measurement precision
If mobile devices use traditional navigation methods, then navigation accuracy can be achieved, but use of energy increases due to prolonged computation time
Solution Approach 1:
The patent segments the computationally intensive particle filter algorithm into multiple parallel clone processes that can be executed more efficiently. By dividing the computational task into manageable segments running in parallel, the system reduces the total computation time and consequently lowers power consumption, while still achieving the same navigation accuracy through the combined results of the clone processes.
Solution Approach 2:
The system applies discarding and recovering by terminating clone processes that prove incorrect based on motion data, while preserving and building upon the correct hypothesis. This allows the system to discard computationally expensive incorrect paths early and recover computational resources to confirm the correct heading, reducing overall energy consumption while maintaining accuracy.
Data Source
AI summary
Estimating initial heading at start-up of navigation. At least some of the example embodiments are computer-implemented methods including: spawning a plurality of clone processes, each clone process given an identical location and speed, and each clone process given a unique direction; calculating, by each clone process, a respective position at the end of a frame period; terminating clone processes whose position at the end of the frame period is outside a predetermined threshold, the terminating results in remaining clone processes; and determining the heading of the mobile device from the remaining clone processes.