After the datacenter produces the public and individual messages, these messages must be published to the right targets. We divide the targets into three categories: individual, public, and open platform for the emergency network. The backend cloud system sends unique information to the different targets.Individual emergency plans include two functions for smartphone users: routing guide and remote telehealth. The remote telehealth service focuses on information representation and synchronous tasks on distributed datacenters. We propose a hidden Markov predicting model (HMPM)
for predicting bandwidth. This model arouses the system more quickly than does increasing the number of requests.10
The information posting system also includes a public alarm system and public emergency policies.
The public alarm system triggers smartphones to alert users in an affected area when the cloud-based system predicts a disaster. Users can customize the smartphone alarm app to receive different types of alarms, such as floods, volcanos, and tsunamis. Public emergency policies allow the posting of disaster conditions, which are also posted through open plat-
forms, such as an official site, social media services, and radio and television broadcasts.