Multicast and data broadcast are efficient approaches for 
one-to-many data delivery. In this talk, we first focus on the 
allocation of forwarding states for IP multicast. Our approach
leverages an emerging IETF standard, Explicit Multi-Unicast
(Xcast), and our algorithm obtains the optimal assignment of
 multicast forwarding states. Compared with previous works, our
approach leads to much fewer states stored in each multicast tree,
and the allocation of forwarding states is more balanced. 
Afterward, we propose a new data broadcast framework based on 
network coding. The existing data broadcast mechanisms deliver each
dedicated queried data item directly. In contrast, we first 
allocate and mix multiple queried data items with network coding
and then send the coding data. One advantage of our algorithm is 
that the number of delivered data items is much smaller than the
number of queried data items. In addition, our algorithm leads to 
the minimum decoding costs in clients.