With the rapid urbanization of China, the urban developing land is spreading outward. The phenomenon of separation between housing and employment is becoming increasingly serious. Long commuting brings physical and mental harm and social problems. The analysis of job-housing relationship is conducive to the rational allocation of work and residence space, reducing urban commuting time and costs, improving the efficient utilization of urban space, and promoting high-quality urban development. The previous research on the relationship between employment and housing mainly focused on census data and questionnaire survey data. With the application and development of big data, new types of data such as bus swiping data (SCD), social media data, Baidu Huiyan data, and mobile signaling data are gradually being applied. However, these studies often analyze the job-housing relationship at a specific time, lacking analysis of the dynamic changes. Chongqing, a typical cluster city, is located in the southwest of inland China. During the rush hour, there is a high demand for traffic and severe congestion in Chongqing, which seriously affects the urban efficiency and the residents’ health. Based on the mobile signaling data, this paper analyzes the changes in job-housing relationship in the central urban area of Chongqing in 2015 and 2021. The mobile signaling data is provided by Smart Footprint Data Technology Co., Ltd. after desensitizing and expanding the data generated by all Unicom mobile phone users in the central urban area of Chongqing. These data are composed of 250 m * 250 m grids, and are divided into working population data, residential population data, and the average origin to destination (OD) data of one month. Based on the calculation of the kernel density, the ratio of occupation to housing, the employment self-sufficiency, with referring to the urban spatial pattern of the “Chongqing Urban Rural Master Plan (2007-2020)”, this paper compares the job-housing relationship in the central urban area of Chongqing from two levels. The results show that: 1) With the development of cities and urban expansion, the phenomenon of separation of occupation and residence is inevitable. From 2015 to 2021, 21 clusters in the central urban area of Chongqing have significant differentiation in terms of job and housing, and shows a trend of northward development, which is consistent with the urban functional layout in the overall urban planning. 2) There is a basic balance between employment and housing in the central urban area of Chongqing in 2015 and 2021. A comparative analysis of job-housing relationship in five major regions shows that there are significant changes in employment and housing between the eastern and southern regions of Chongqing. The employment and housing in the northern part of the eastern region tends towards employment, while the employment and housing in the southern and southern parts of the eastern region tends towards residence. In addition, a comparative analysis of 21 clusters indicates that there is a significant polarization in the relationship among clusters, with an overall trend of northward development. And the decrease in self-sufficiency is significant, and residents’ employment is no longer limited in a small range, resulting in an increase in the proportion of commuting among different clusters. 3) In the context of the new round of urban and rural planning, various clusters will be integrated and reclassified, and the job residence relationship analysis will be conducted based on 9 clusters. The results show that the functions of Yuzhong, Nanpeng, Taojia, and Longsheng clusters are biased towards employment, the functions of the Chayuan, Beibei, and Diaoyuzui-Longzhouwan clusters are biased towards residence, the functions of the Liangjiang and Jinfeng clusters are balanced between job and residence. The commuting efficiency of Nanpeng and Longsheng is relatively low. The matching degree between job and housing in Chayuan cluster is relatively low, as well as the commuting efficiency. There are still some shortcomings in this study, and further research is urgently needed. As the mobile signaling data obtained in this study only have O (starting point) and D (ending point), there is a lack of path expression, and cannot form a complete travel chain. As a result, the traffic flow cannot be loaded onto the road and cannot quantitatively evaluate the job-housing and the commuting. In the future, the specific reasons for congestion and the factors affecting residents’ excessive commuting would be explored.