Human mobility data are fundamental to Web-based location services and broader urban applications, and mobility generation provides synthetic trajectories when real data are costly or inaccessible. Although recent mobility generators have incorporated collective mobility patterns, they primarily focus on individual trajectories and aggregate population dynamics, leaving evolving community structures formed by user-location interactions insufficiently modeled. In this paper, we study community-consistent human mobility generation. To this end, we formulate mobility generation as the generation of dynamic user-location graphs to model community structures and propose MOTION, a Multi-agent framework for cOmmuniTy-consIstent human mObility geNeration. MOTION consists of user agents, location agents, and a global manager agent. MOTION first detects coherent communities from historical mobility using the Leiden algorithm and extracts community-conditioned interaction priors to guide the generation process. Guided by these priors, the user agents plan mobility, search destination candidates, and infer user movements from the individual view. In parallel, selectively activated location agents infer likely user arrivals from the community view. The manager coordinates both views, diagnoses interactions causing deviations, and attributes targeted reflection signals to revise user decisions, location attraction, and source-user selection. Experiments on two real-world datasets show that MOTION outperforms baselines by at least 4.05% across reported metrics of individual trajectories, population and flow distributions, and community and graph structures. Further analyses confirm its robustness, privacy preservation, and downstream utility. The code is available at https://anonymous.4open.science/r/MOTION-04C4.
Figures & tables
Figure 1. Human mobility’s three layer patterns and corresponding properties and downstream applications.
Figure 2. Community structures detected from real and generated mobility. Different colors represent different communities on the same gridded map.
Figure 3. Overview of MOTION. MOTION first extracts community structures and interaction priors from real-world mobility. During generation, the user agent and the location agent propose user-location graphs from the individual and receiver-side community views. The manager coordinates these proposals and uses the extracted priors to identify structural deviations and guide targeted reflection.
Figure 4. Local-set formation. Locations and their neighborhoods are activated according to population dynamics, nearby moving users, and prior-supported incoming flows and are then merged into a local set.
Figure 5. Community-consistency reflection. MOTION bridges generated and real communities, diagnoses deviations, attributes them to related decisions, and sends reflection signals to regenerate the affected proposals.
Pattern Layer
Evaluation Metrics
Individual Trajectory Statistics
Distance, Radius, LocNum
Population and Flow Distributions
Pop-EMD, Flow-EMD, G-Rank
Community and Graph- Structural Consistency
Comm-Size, Comm-Mem, Motif, W-Degree
Table 1. Evaluation metrics of three pattern layers.
Methods
Individual Trajectory Statistics
Population and Flow Distributions
Community and Graph-Structural Consistency
Distance ↓
Radius ↓
LocNum ↓
Pop-EMD ↓
Flow-EMD ↓
G-Rank ↓
Comm-Size ↓
Comm-Mem ↑
Motif ↓
W-Degree ↓
Private Car Dataset
DRTISA
0.2645 ±0.0038
0.1951 ±0.0031
0.1518 ±0.0027
0.0534 ±0.0011
0.0511 ±0.0012
0.2741 ±0.0042
0.0876 ±0.0015
0.2661 ±0.0033
0.2744 ±0.0041
0.3637 ±0.0050
SeqGAN
0.2680 ±0.0040
0.2050 ±0.0035
0.1896 ±0.0030
0.0597 ±0.0013
0.0590 ±0.0014
0.2889 ±0.0045
0.0725 ±0.0014
0.4388 ±0.0078
0.1517 ±0.0024
0.2898 ±0.0042
MoveSim
0.2257 ±0.0034
0.1781 ±0.0029
0.2575 ±0.0038
0.1037 ±0.0020
0.1014 ±0.0019
0.3475 ±0.0051
0.2203 ±0.0035
0.2756 ±0.0033
0.2918 ±0.0044
0.3134 ±0.0047
DeepMobility
0.2086 ±0.0053
0.1185 ±0.0019
0.2190 ±0.0032
0.0565 ±0.0018
0.0555 ±0.0024
0.2614 ±0.0030
0.0977 ±0.0012
0.5335 ±0.0041
0.2263 ±0.0038
0.3007 ±0.0044
Table 2. Overall performance on the private car and mobile phone datasets (mean ± standard deviation over five runs where available; otherwise, mean only). Higher Comm-Mem and lower values of all other metrics indicate better performance. Bold indicates the best result, and underline denotes the second-best.
Variants
Dist. ↓
Radius ↓
Pop-EMD ↓
Comm-Size ↓
Comm-Mem ↑
Private Car Dataset
w/o user
0.1939
0.1918
0.0459
0.0887
0.4868
w/o location
0.1817
0.1050
0.0524
0.0881
0.5320
w/o manager
0.2054
0.1331
0.0486
0.0831
0.4677
single-agent
0.2367
0.1174
0.0548
0.0792
0.5247
deterministic
0.2511
0.1219
0.0665
0.0841
0.4925
Table 3. Ablation results on both datasets. The single-agent and deterministic controls use the same information as MOTION. Bold indicates the best result.
Figure 6. Community structures detected from real and generated mobility on the private car and mobile phone datasets. The panels compare real mobility and MOTION with CoDiffMob on the private car dataset and LLMob on the mobile phone dataset; each color distinguishes a detected community.
Figure 7. Impact of different LLM backbones on the private car and mobile phone datasets.
Figure 8. Sensitivity of Distance (left) and Pop-EMD (right) to top_k and active_m on the private car and mobile phone datasets.
Figure 9. An example of multi-agent reasoning. Colored regions highlight the key inputs and outputs of each agent role.
Figure 10. Privacy results: (a) nearest-training-trajectory similarities on the private car dataset and (b) membership inference attack success rates on both datasets.
Figure 11. Downstream prediction performance on the private car dataset under the 100% (left) and 50% (right) training settings. Higher values indicate better performance.
Methods
Private Car Dataset
Mobile Phone Dataset
Tokens
Latency
Tokens
Latency
MOTION
40,662
229 s
43,132
224 s
LLMob
40,647
255 s
45,023
230 s
AgentMove
42,331
264 s
48,159
244 s
Table 4. End-to-end token consumption and latency per generated user-day.
Figure 12. Downstream prediction performance on the private car dataset under the 100% (left) and 50% (right) training settings. Higher values indicate better performance.
Methods
Private Car Dataset
Mobile Phone Dataset
Tokens
Latency
Tokens
Latency
MOTION
40,662
229 s
43,132
224 s
LLMob
40,647
255 s
45,023
230 s
AgentMove
42,331
264 s
48,159
244 s
Table 5. End-to-end token consumption and latency per generated user-day.
Dataset
Period
# Locations
# Users
# POIs
# POI Categories
# Records
Avg. Moves per User
Avg. Daily Moves per User
Private car dataset
2 weeks
500
2,342
717,461
14
678,341
8.75
0.88
Mobile phone dataset
2 weeks
500
5,352
679,413
14
1,235,619
18.60
1.28
Table 6. Statistics of the private car and mobile phone datasets.
Appendix figures & tables11 assets
Supplementary material from the paper’s appendix.
Appendix
POI Semantic Categories
Transportation Facilities
Leisure & Entertainment
Companies & Enterprises
Healthcare
Commercial & Residential
Tourist Attractions
Automotive
Life Services
Science & Education & Culture
Shopping & Consumer Goods
Sports & Fitness
Hotels & Accommodations
Appendix
Table 7. The 14 high-level POI semantic categories.
Motif
Representative Examples
STAY
Commercial & Residential (no movement); Companies & Enterprises (no movement).
Table 8. Representative mobility motif examples using POI semantics and discretized travel distances.
Figure 13. Cumulative distributions of the Top-1, Top-2, and Top-3 similarities between generated trajectories and their nearest real trajectories in the training split of the mobile phone dataset.
Figure 14. Impact of the maximum number of within-step reflection rounds on the private car and mobile phone datasets. Lower values indicate better performance.
Figure 15. Cross-dataset generalization between the private car dataset (P) and mobile phone dataset (M). Rows indicate the source dataset, while columns indicate the target dataset.
Component
Private Car Dataset
Mobile Phone Dataset
Tokens
Latency
Tokens
Latency
user agent
436.1
1.57 s
372.3
2.21 s
location agent
1,345.0
2.15 s
1,462.0
1.98 s
coordination
720.5
2.07 s
806.3
2.06 s
reflection
866.4
3.77 s
981.0
3.11 s
Appendix
Table 9. Component-level token consumption and latency per reasoning call.
Figure 16. Additional sensitivity results of top_k and active_m on the private car and mobile phone datasets. Lower values indicate better performance.
Figure 17. Prompt template for individual-view destination reasoning by a user agent.
Figure 18. Prompt template for receiver-side community reasoning by a location agent.
Figure 19. Prompt template for individual-community coordination by the manager agent.
Figure 20. Prompt template for community-consistency reflection by the manager agent.