Efforts to integrate sensor networks into internet have been widely pursued. A FTR-HTTP (Finite Time Response Hyper Text Transfer Protocol)has beenproposed to integrate the sensor networks witha data collector server through internet. The FTR-HTTP has two controlvariables, i.e. forced sleep duration (K) and maximum response time (MaxTr),in which both affect the performance of the FTR-HTTP. This paper reports some experimental researches that demonstratethe effects of variations of the two control variables on the performance of the FTR-HTTP. The results show that increasing the value of Kvariable has an effect on the improvements of three performance measures of FTR-HTTP, i.e. average response time, average energyefficiency, and average request timeout. Where as, increasing the value of Kvariable can increase the value of average update time, which can be described as a sampling frequency of data transmission to the server. It means that increasing the value of Kvariable can improve thethree performance measures (average response time, average energy efficiency, and average request timeout), but trade off must be made in the expense of averageupdate time. The results of the experiment show that increasing the value of Max TR variable has little positive effect on three performance measures of the FTR-HTTP (average response time, energy efficiency, and average update), but it has little negative effects on the other performance measure of the FTR-HTTP (average request timeout). Finally, a closed-loop control system being able to achieve the optimal performance of the FTR-HTTP, by controlling the value of K and Max TR variables, can be proposed for futureworks.